StorageWorks™ Array Controllers HS Family of Array Controllers User’s Guide Order Number: EK–HSFAM–UG. D01 The StorageWorks Array Controllers HS Family of Array Controllers User’s Guide contains instructions for installing and using HSJ30, HSJ40, HSD30, HSZ40–Ax, and HSZ40–Bx array controllers.
March 1995 The information in this document is subject to change without notice and should not be construed as a commitment by Digital Equipment Corporation. Digital Equipment Corporation assumes no responsibility for any errors that may appear in this document. The software described in this document is furnished under a license and may be used or copied only in accordance with the terms of such license.
Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv 1 Introduction to HS Array Controllers 1.1 1.2 1.3 1.3.1 1.3.2 1.3.3 1.4 1.5 1.5.1 1.5.2 1.5.3 1.6 1.7 1.8 1.9 1.10 1.10.1 1.10.2 1.11 1.12 1.13 Overview of HS Array Controllers . . . . . . . . . . . . . . . . . . . . . . Housing for HS Array Controllers . . . . . . . . . . . . . . . . . . . . . . Physical Description of HS Array Controllers . . . . . . . . . . . . .
2.1.11 Dual Controller Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 HS Array Controller Firmware Overview . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Core Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1.1 Controller Self-Test and Diagnostics . . . . . . . . . . . . . . . . . . . . 2.2.1.2 Executive Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Host Interconnect Functions . . . . . . .
4 Installation 4.1 Customer Site Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Power and Power Cord Requirements . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.2 Shelf Power Configuration Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.3 Environmental Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Before You Begin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.6 Power Supply Status LEDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 Battery Backup Unit (BBU) Status LEDs . . . . . . . . . . . . . . . . . . . . . . . . . 5.8 Environmental Monitor Unit (EMU) for HSZ40–Bx Array Controllers . . . 5.8.1 EMU Fault Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.8.2 Controller Fault Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.9 Description of Device Warm Swap . . .
6.10.1 Configuring Host Units into Mirrorsets . . . . . . . . . . . . . 6.11 Mirrorset Command Overview . . . . . . . . . . . . . . . . . . . . . . . 6.11.1 Creating a Mirrorset . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11.2 Mirrorset SET Commands and Qualifiers . . . . . . . . . . . 6.11.3 Mirrorset SHOW Commands . . . . . . . . . . . . . . . . . . . . . 6.11.4 REDUCE disk-device-name1 Command . . . . . . . . . . . . . 6.11.5 MIRROR disk-device-name container-name Command . . 6.11.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–59 7–60 7–60 7–60 7–62 7–62 7–63 7–64 7–65 7–66 7–67 7–68 7–68 7–71 7–71 7–71 7–72 7–72 7–74 7–75 7–75 7–85 7–101 7–101 7–101 7–101 7–102 7–103 7–103 7–103 8.1 Digital Supported Operating Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 HS Array Controller System Management . . . . . . . . . . . . . . . . . . . . . . . . . 8.3 OpenVMS and VMS VAX Operating Systems . . . . . . . . . . . . . . . . . . . . . . 8.3.
8.4.1 uerf with DEC OSF/1 for HSZ Array Controllers . . . . . . . . . . . . . . . . . 8.4.2 DECsafe Available Server Environment (ASE) . . . . . . . . . . . . . . . . . . 8.4.3 Configurations and Device Support for the HSZ Array Controllers . . . 8.4.3.1 Virtual Terminal Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.3.2 DEC OSF/1 Device Special Files for HSZ Array Controllers . . . . . 8.5 Basic Steps for Configuring an HSZ Array Controller Under the DEC OSF/1 Operating System .
HELP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . INITIALIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LOCATE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MIRROR disk-device-name1 container-name . . . . REDUCE disk-device-name1 [disk-device-nameN] RENAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RESTART OTHER_CONTROLLER . . . . . . . . . . . RESTART THIS_CONTROLLER . . . . . . . . . . . . . RETRY_ERRORS UNWRITEABLE_DATA . . . . .
B.2 B.2.1 B.2.2 B.2.3 B.2.4 B.3 SHUTDOWN THIS_CONTROLLER UNMIRROR disk-device-name . . . . . CLI Messages . . . . . . . . . . . . . . . . . . Error Conventions . . . . . . . . . . . CLI Error Messages . . . . . . . . . . Warning Conventions . . . . . . . . . CLI Warning Messages . . . . . . . Device Configuration Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B–4 B–5 B–6 B–7 B–8 B–9 B–10 B–11 B–12 B–13 B–14 Setting the Terminal Speed and Parity . . . . . . . . . . . . . . . . . . Creating a Unit from a Disk Device . . . . . . . . . . . . . . . . . . . . . Creating a Unit from a Tape Device . . . . . . . . . . . . . . . . . . . . . Creating a Unit from a Four-Member Stripeset . . . . . . . . . . . . Creating a Unit from a Five-Member RAIDset . . . . . . . . . . . . Creating a Unit from a Disk Device and Setting the Write Protection . . . . . . . . . . . . . . .
3–8 3–9 3–10 3–11 3–12 4–1 4–2 4–3 4–4 4–5 4–6 4–7 4–8 4–9 4–10 4–11 4–12 5–1 5–2 5–3 5–4 5–5 5–6 5–7 5–8 5–9 5–10 7–1 7–2 7–3 7–4 7–5 7–6 7–7 7–8 7–9 7–10 7–11 7–12 7–13 7–14 7–15 7–16 7–17 7–18 Adjacent Devices on a Single Port . . . . . . . . . . . . . . . . . . . . . . . . . . SW300-Series Cabinet Controller/Storage Shelf . . . . . . . . . . . . . . . . Balanced Devices Within Storage Shelves . . . . . . . . . . . . . . . . . . . . Optimal Availability Configuration Example . . . . . . . . . . . . . .
Tables 1 1–1 1–2 1–3 1–4 3–1 3–2 3–3 3–4 3–5 3–6 3–7 3–8 3–9 4–1 4–2 4–3 4–4 4–5 4–6 5–1 5–2 5–3 5–4 5–5 7–1 7–2 7–3 8–1 A–1 A–2 A–3 A–4 A–5 A–6 A–7 A–8 xiv HS Array Controller Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of HS Array Controller Features . . . . . . . . . . . . . . . . . . . . . HSJ Array Controller and Cache Module Specifications . . . . . . . . . . . HSD30 Array Controller and Cache Module Specifications . . . . . . . . .
Preface Introduction This StorageWorks Array Controllers HS Family of Array Controllers User’s Guide contains instructions for installing and using HSJ30, HSJ40, HSD30, HSZ40–Ax, and HSZ40–Bx array controllers. For purposes of this manual, the term ‘‘HS array controller’’ refers to several models, as shown in Table 1.
Chapter 3 Describes the configuration rules and restrictions for standard and nonstandard HS array controller subsystems. The maximum lengths for host port cables is listed, as well as supported host adapters for each controller model.
Document Title Order Number HSD30 Firmware Array Controller Software Product Description (SPD 53.53.03) for V2.5 HSOF AE–Q6HKD–TE HSZ40 Firmware Array Controller Software Product Description (SPD 53.54.04) for V2.5 HSOF AE–Q6HME–TE HS Family of Array Controllers Optional Operating Firmware (HSOF) Software Product Description (SPD 53.38.01) for V2.5 AE–QDA4B–TE StorageWorks Array Controllers HS Family of Array Controllers Service Manual (Use revision D01 for HSOF Version 2.5.
Documentation Conventions The following conventions are used in this manual: xviii boldface type Boldface type in examples indicates user input. Boldface type in text indicates the first instance of terms defined in either the text, in the glossary, or both. Boldface type is used also for controller model names to help the reader pick out the information for their specific model in procedures that talk about the differences of each model.
1 Introduction to HS Array Controllers This chapter contains an overview of the HS family of array controllers. A list of key attributes for each controller and the controller shelves are provided. This chapter also provides a brief description of the protocols used with each controller model’s functional environment. 1.
Note HSJ30 and HSJ40 array controllers hereafter are referred to as HSJ array controllers unless the text is controller-model specific. HSZ40–Ax and HSZ40–Bx array controllers hereafter are referred to as HSZ array controllers unless the text is controller-model specific. 1.2 Housing for HS Array Controllers The controller modules and associated read and write-back cache modules mount in a controller shelf.
availability options also are sacrificed. Digital does not recommend this configuration. • The six SCSI–2 device cables (three for HSJ30) from the controller shelf can be routed to more than one SBB shelf as desired, taking into consideration prescribed configuration rules and restrictions as listed in Chapter 3.
Figure 1–1 HS Array Controller Subsystem in an SW800-Series Data Center Cabinet CONTROLLERS 800-SERIES CABINET DEVICE CXO-4114A-MC Figure 1–2 depicts a block diagram of a dual-redundant HSJ30 controller subsystem. In an HSJ30 dual-redundant configuration, two controllers attach to three SCSI–2 buses within the same controller shelf. Note The same block diagram applies to the HSJ40 controller, except the HSJ40 controller has six SCSI–2 buses. 1.3.
Figure 1–2 Dual-Redundant HSJ30 Array Controller Subsystem—Block Diagram HOST CPU CI CI MAINTENANCE TERMINAL EIA423 PORT HSJ30 CONTROLLER (OPTIONAL) 3 2 1 CI STAR COUPLER READ OR WRITEBACK CACHE READ OR WRITEBACK CACHE FAILOVER COMMUNICATION HSJ30 CONTROLLER MAINTENANCE EIA- TERMINAL 423 PORT (OPTIONAL) 1 2 3 CXO-3699D-MC • Each SCSI–2 port supports up to six (or seven) SCSI–2 devices,3 , depending on the subsystem configuration.
Supported configurations allow up to four HSD30 array controllers per DSSI bus (for example, four nonredundant controllers, or two dual-redundant controller pairs, or one dual-redundant controller pair and two nonredundant controllers). Refer to the StorageWorks Array Controllers HSD30 Array Controller Operating Firmware Release Notes for changes and/or restrictions. Note Separate buses are supported for an HSD30 array controller dualredundant configuration using the OpenVMS Version 6.
• The six SCSI–2 port cables from the controller shelf can be routed to more than one device shelf as desired, taking into consideration prescribed configuration rules and restrictions as listed in Chapter 3. The HSZ array controller consists of a single module that is configured alone (nonredundant) or with a second like controller for dual-redundant availability in the controller shelf. One read cache or write-back cache module is supported for each controller module.
1.5 Addressing Storage Within the Subsystem This section provides an overview of how a controller subsystem addresses storage. Storage is seen in two different ways, depending on your perspective and your controller model: 1. From the controller SCSI–2 device interface—At the physical device level 2. From the host interface—At the virtual device level The following sections describe both levels of storage addressing. 1.5.
Figure 1–3 Controller Storage Addressing HOST INTERFACE CONTROLLER DEVICE INTERFACE PORT 1 PORT 2 PORT 3 PORT 4 PORT 5 CONTROLLER PORT ADDRESS PORT 6 SCSI BUS 1 SCSI BUS 2 SCSI BUS 6 SCSI BUS 3 SCSI BUS 4 SCSI BUS 5 (OPTIONAL) SCSI ID SCSI ID SCSI ID 5 4 6 LUN 0 LUN 0 SCSI ID 3 SCSI ID 2 SCSI ID 1 SCSI ID 0 CONTROLLER TARGET ADDRESS LUN 0 LUN 0 LUN 0 LUN 0 LUN 0 CONTROLLER LUN ADDRESS DEVICE 4 DEVICE 3 DEVICE 2 DEVICE 1 DEVICE 0 LUN 1 DEVICE 6 DEVICE 5 StorageWorks STORAGE S
of storing data, whether it is a physical device or a group of physical devices. A disk, a stripeset, a RAIDset, or a mirrorset are examples of containers. Controller Port Target LUN Addressing Controller Port Target LUN (PTL) addressing is the process by which the controller selects storage space within a specific, physical, storage device. The process takes place in three steps: 1. The port selection—The controller selects the SCSI–2 bus port connected to a particular device. 2.
1.5.3 Host Storage Addressing (HSZ Array Controllers) Figure 1–4 shows a typical connection between an HSZ array controller and its host. In this case, the SCSI–2 host device interface consists of device ports (buses), each connected to a SCSI–2 bus containing up to eight devices. The HSZ array controller resides on one of the SCSI–2 buses.
Controller LUNs and SCSI host LUNs may represent the same structure, but only if the user configures (up to) eight controller devices in a one-toone unit relationship with the host. This situation rarely occurs under normal operation. Host Port Target LUN Addressing (HSZ Array Controllers) Note Non-SCSI hosts (CI, DSSI), though they access virtual devices, do not use a PTL addressing scheme. Any unit seen by these hosts is simply called a host logical unit (not a LUN).
• HSZ array controllers can use a SCSI virtual terminal connection from a host application utility called HSZterm (or StorageWorks HSZ40 Array Controller Utility for DEC OSF/1 Version 1.0). Refer to the StorageWorks Configuration Manager for DEC OSF/1 Installation Guide and the StorageWorks Manager for DEC OSF/1 System Manager’s Guide for HSZterm for complete HSZterm installation and operating instructions.
Table 1–1 (Cont.
Table 1–3 HSD30 Array Controller and Cache Module Specifications Specification Controller Module Typical Controller Module Maximum Read Cache 16 MB Read Cache 32 MB Write-Back Cache 16 MB Write-Back Cache 32 MB Width 12.5 inches N/A 12.5 inches 12.5 inches 12.5 inches 12.5 inches Depth 8.75 inches N/A 7.75 inches 7.75 inches 7.75 inches 7.75 inches Power Consumption 20.76 watts 23.04 watts 1.82 watts 2.02 watts 2.28 watts 2.48 watts @ +5 V 4.15 A 4.
1.10 StorageWorks Controller Subsystem Products Digital offers several preconfigured HS array controller subsystems. However, if you wish to customize your own system, read the StorageWorks Solution Configuration Guide and the StorageWorks Solutions Shelf and SBB User’s Guide for guidance in creating a customized, configure-to-order (CTO) subsystem.
Figure 1–6 5¼-Inch Storage SBB DEVICE ACTIVITY (GREEN) DEVICE FAULT (AMBER) CXO-4308A-PH A 3½-inch SBB occupies one slot in a BA350–SB SBB shelf. Three slots are required to house 5¼-inch SBBs. The currently supported cabinets that house HS array controllers, options, and devices are the SW800-series data center cabinets, the SW500-series cabinets, or the SW300-series deskside RAID enclosure cabinets.
Figure 1–7 Controller Shelf with a Dual-Redundant HSJ40 Array Controller Configuration CXO-3660A-PH Note In a dual-redundant controller configuration with the controller shelf mounted vertically in the front of an SW800-series data center cabinet, or horizontally in a SW500-series cabinet, the controller closest to the SCSI– 2 device cables is ID #6; the one farthest from the device cables is ID #7.
Figure 1–8 BA350–SB Fully Populated SBB Shelf SLOT 7 (POWER SUPPLY ONLY) SLOT 6 (POWER SLOT OR 5 SLOT I/O DEVICE) 4 SLOT 3 SLOT 2 SLOT I/O 1 SLOT DEVICES 0 ONLY SCSI BUS CONNECTORS EMI GASKET SCSI CABLE ACCESS CXO-4231A-MC 1.11 MSCP and TMSCP Protocols (to Hosts) The CI and DSSI interfaces use MSCP and TMSCP as the storage protocols to access HSJ and HSD30 controller-attached SCSI–2 devices. MSCP and TMSCP are high-level storage protocols used by many Digital storage subsystem products. 1.
2 Controller Technical Description The following sections present brief overviews of HS array controller hardware and firmware functionality. 2.1 HS Array Controller Hardware Functional Overview Each HS array controller provides connections to a host computer and an array of storage devices connected by SCSI–2 buses. A second controller also may be interconnected in a redundant arrangement called a dual-redundant configuration, where both controllers reside in the same controller shelf.
Figure 2–1 HS Array Controller Functional Block Diagram OCP POLICY PROCESSOR 32KB I/D CACHE INTEL 80960CA/CF µP DIAGNOSTIC REGISTERS NVMEM RAID ASSIST DEVICE PORT 1 DEVICE PORT 2 MAINTENANCE TERMINAL PORT DUAL CONTROLLER PORT 16 MB OR 32 MB READ OR WRITE-BACK CACHE (OPTION) BUS EXCHANGER 8 MB SHARED MEMORY HOST INTERFACE PROGRAM CARD DEVICE PORT 3 DEVICE PORT 4 DEVICE PORT 5 DEVICE PORT 6 CXO-4178D-MC 2.1.
2.1.3 CI, DSSI, or SCSI–2 Interfaces (Host Ports) A CI, DSSI, or SCSI–2 interface allows direct memory access of data between the host port and shared memory. Setup and maintenance of the host port is done by the policy processor hardware. 2.1.4 SCSI–2 Device Ports (Buses) The SCSI–2 device ports for the controllers are implemented using SCSI–2 port processor chips performing 8-bit operations in normal or FAST mode.
the impact of power failure is worsened by the possibility of write hole data loss as well. For this reason, the write-back cache differs from the read cache by incorporating onboard rechargeable batteries. The batteries power the memory (to retain data) when cache power is intentionally or accidentally interrupted. The battery circuit automatically detects loss of power and switches from shelf backplane power to battery power. 2.1.6.
When possible, single cache battery failures will cause HSJ- and HSD30-based RAIDsets and mirrorsets to fail over to the companion cache in a dual-redundant configuration provided the other cache’s batteries are fully charged.2 Although this feature will not help during power outages (because both controllers will be down), it keeps your RAIDsets and mirrorsets online if, for example, one battery is faulty.
SHOW OTHER_CONTROLLER command to check the status of your batteries. See Section 4.10 for information on how to upgrade your cache module option from a read to write-back cache module (or from 16-MB to 32-MB write-back cache module). The battery circuit automatically detects loss of backplane power and switches from backplane +5V power to battery power. This transition occurs when the backplane supply drops to approximately +4.65V +/- .
Figure 2–2 HSJ40 Array Controller OCP and Program Card Locations PROGRAM CARD HSJ40 RESET BUTTON OCP CXO-4335A-MC The OCP for the HSJ40 array controller has seven square plastic buttons with embedded LEDs. The HSJ30 array controller uses the same OCP, but only three of the six port buttons can be used functionally to quiesce a port. All six amber LEDs on the HSJ30 and HSJ40 array controller’s OCPs report fault codes. Both controllers utilize the reset button. The buttons and LEDs serve two functions.
The buttons and LEDs serve two functions. The green button resets/restarts the controller. The three recessed buttons allow the port (associated with each button) to be quiesced when the button is pushed, allowing for a device warm swap, thus decreasing subsystem down time. Use a pointed object to push the three recessed port buttons. Figure 2–3 shows an HSD30 array controller OCP. Figure 2–3 HSD30 Array Controller Operator Control Panel RESET BUTTON OCP HSD30 PORT LEDS PORT BUTTONS CXO-4360A-MC 2.1.
Figure 2–4 HSZ40–Bx Array Controller Operator Control Panel RESET BUTTON HSZ40B 1 2 3 4 5 6 HSZ40B 123456 PORT LEDS OCP PORT BUTTONS PCMCIA PROGRAM CARD SLOT PROGRAM CARD EJECT BUTTON CXO-4204C-MC The buttons and LEDs serve two functions. The green button resets/restarts the controller. The six recessed buttons allow the port (associated with each button) to be quiesced when the button is pushed, allowing for a device warm swap, thus decreasing subsystem down time.
The program card must remain inserted with the ESD shield in place, whenever the controller is in operation. Removing the program card during operation causes the controller to go into a reset state. To recover from this state, push the program card back into place and press the controller’s reset (//) button. Figure 2–5 shows the program card location in a controller front bezel and its associated eject button.
Note For HSZ40–Bx array controllers, the MMJ connector is center-keyed (rather than offset-keyed) and ships with a short phone cable (Digital part number 17–03411–04) and a center-to-offset coupler (Digital part number 12–43346–01). These parts are used to convert from a center-keyed to offset-keyed maintenance terminal connector configuration. A terminal is plugged into the MMJ connector during subsystem installation to set initial controller parameters.
See Figure 2–7 for a visual representation of the HSZ40–Bx array controller’s short phone cable and center-to-offset key adapter. Figure 2–7 Phone Cable and Center-to-Offset Coupler for HSZ40–Bx Array Controller MMJ Maintenance Terminal Port HSZ40B 123456 CENTER KEYED CENTER KEYED SIDE VIEW OFFSET KEYED FRONT VIEW CXO-4338B-MC 2.1.
• Operator interface and subsystem management functions—command line interpreter (CLI), DUP support for HSJ and HSD30 array controllers, and HSZterm support for HSZ array controllers • Local programs • Error logging and fault management • Device services functions—SCSI–2 port control firmware • Value added functions—Data mapping, caching, state change, failover, storageset member management, and error recovery Note This chapter does not attempt to discuss every controller function or provide comple
• Digital Storage System Interconnect (DSSI) for HSD30 Array Controllers System Communication Services (SCS) Mass Storage Control Protocol (MSCP) Tape Mass Storage Control Protocol (TMSCP) Diagnostic Utility Protocol (DUP) • Small Computer System Interface (SCSI) for HSZ Array Controllers SCSI–2 protocol with tagged command queuing and vendor-unique diagnostic pages (for communication with the controller CLI). 2.2.
• CLONE utility for making block-for-block copies of a single device, mirrorset, or striped mirrorset 2.2.3.5 Error Logging and Fault Management Error logging and fault management collects system errors in a central location and sends the error information to the host. 2.2.4 Device Services SCSI–2 device service firmware includes device port drivers and physical device addressing and access algorithms. Device services consist of normal functions such as read and write, as well as error recovery code.
• Connectivity functions: A second controller can be connected to a single controller configuration to allow failover support of attached devices in the event of a controller failure. This is called a dual-redundant configuration. EIA–423 compatible terminals can be connected to each controller’s maintenance terminal port. A VAXcluster™ console system (VCS) or serial interface also can be connected to the maintenance terminal port for maintenance and installation.
In these cases, the command is directed to the correct controller: • THIS_CONTROLLER refers to the controller to which the terminal is connected or which is the target of the virtual terminal connection. • OTHER_CONTROLLER refers to the other controller in the dual-redundant pair. Note Devices, stripesets, RAIDsets, and mirrorsets are defined as containers. All changes in container configuration are automatically communicated between the two controllers. 2.3.
because, after setting failover, you will not be accessing those units anyway. (The target controller will only access the copied units.) See Section 5.2.5 for a detailed description for setting configuration parameters for dual-redundant configurations. HSZ Array Controllers: Observe the following considerations when setting dual-redundant HSZ array controllers for failover: • Subsystem performance will be better if you balance the assignment of target IDs across your dual-redundant pair.
Failover should normally complete in less than 10 seconds. If drive I/O is in progress at the time of failure, the surviving controller must reset any SCSI–2 buses with outstanding I/O. Whenever you need to revive the controller that was disabled, you must enter the following command from a terminal connected to the functioning controller: CLI> RESTART OTHER_CONTROLLER Then, press the reset (//) button to initialize the controller to be revived. 2.3.
2.4 HSZ Array Controller Failover Operation The HSZ array controller uses a transparent failover operation in which the transfer of storage subsystem control occurs in a manner transparent to the host. Two HSZ array controller modules installed in a StorageWorks controller shelf and connected to the same host SCSI–2 bus operate as a redundant pair with transparent failover. Figure 2–8 shows two controllers connected in this manner. Each controller may be configured to have multiple SCSI target IDs.
Figure 2–8 Dual-Redundant Controller Configuration HOST (SINGLE-ENDED INTERFACE) HOST (DIFFERENTIAL INTERFACE) DWZZA SCSI SIGNAL CONVERTER 16-BIT, DIFFERENTIAL INTERFACE REDUNDANT CACHE MODULE CACHE INTERFACE REDUNDANT CONTROLLER MODULE HOST PORT HOST PORT 16 OR 32 MB READ OR WRITE-BACK CACHE MODULE CONTROLLER MODULE ENVIRONMENTAL MONITOR UNIT EIA-423 INTERFACE FAILOVER INTERFACE BACKUP BATTERY (WRITE-BACK CACHE MODULE ONLY) 6 DEVICE PORTS MAINTENANCE TERMINAL OR PC (OPTIONAL) 8-BIT, SINGLE-E
host reinitiates such requests, the surviving controller services them. The time out of the host’s requests is the only indication to the host that a fault of some kind has occurred. The surviving controller continues to monitor the status of the failed controller until it is restarted or replaced by the user.
3 Configuration Rules and Restrictions This chapter describes configuration rules and restrictions for standard and nonstandard (customized) HS array controller subsystems. When specific rules and restrictions are not provided, references are given to the proper StorageWorks documentation. Note Configuration rules and restrictions apply to all HS array controller platforms (HSJ30, HSJ40, HSD30, HSZ40–Ax, and HSZ40–Bx array controllers) unless stated otherwise. 3.
3.2.1 SW800-Series Data Center Cabinet This section describes the rules that apply to controller subsystem configurations in SW800-series data center cabinets. Figure 3–1 shows the loading sequence for storage and controller shelves (without tape drives) in an SW800-series data center cabinet. Refer to the StorageWorks SW800-Series Data Center Cabinet Installation and User’s Guide for new loading sequence illustrations for cabinets using 4 GB or larger disk drives.
• • Two storage shelves per port (jumpered-pairs)—Two BA350–SB shelves can be joined on the same controller port with the following restrictions: – The SCSI–2 cable to the first BA350–SB storage shelf is 1.0 meter or less. The associated BA350–MA controller shelf must be near enough to satisfy this restriction. – The SCSI–2 cable from the first BA350–SB shelf to the second shelf is 0.5 meters or less. This requires two shelves to be immediately adjacent to each other.
Figure 3–2 SW800-Series Data Center Cabinet Loading Sequence with Two Tape Drive Positions SHELF MOUNTING LOCATIONS HOLE #14 TAPE POSITION T2 (C1-C4) TAPE POSITION T1 (C1-C4) STORAGE POSITION S13 (C3) CONTROLLER POSITION C4 HOLE #3 CONTROLLER POSITION C3 SHELF MOUNTING LOCATIONS TAPE POSITION T1 (C1-C4) TAPE POSITION T2 (C1-C4) (C3/C4) STORAGE POSITION S14 CONTROLLER POSITION C2 HOLE #26 HOLE #14 CABLE PASSTHROUGH CONTROLLER POSITON C1 CABLE PASSTHROUGH HOLE #3 HOLE #26 (C2) STORAGE PO
Figure 3–3 SW800-Series Data Center Cabinet Loading Sequence with Four Tape Drive Positions SHELF/TAPE MOUNTING LOCATIONS TAPE (C1-C4) TAPE (C1-C4) POSITION T4 POSITION T3 HOLE #32 TAPE POSITION T1 (C1-C4) TAPE POSITION T2 (C1-C4) TAPE (C1-C4) TAPE (C1-C4) POSITION T3 POSITION T4 CONTROLLER POSITION C2 CABLE PASSTHROUGH TAPE POSITION T1 (C1-C4) CONTROLLER POSITION C1 HOLE #15 TAPE POSITION T2 (C1-C4) CONROLLER POSITION C4 HOLE #3 CONTROLLER POSITION C3 SHELF/TAPE MOUNTING LOCATIONS HOLE #3
Refer to the StorageWorks SW800-Series Data Center Cabinet Installation and User’s Guide for more details. 3.2.2 SW500-Series Cabinets This section describes the rules that apply to controller subsystem configurations in SW500-series cabinets. Figure 3–4 shows the loading sequence for storage and controller shelves in an SW500-series cabinet. Refer to the StorageWorks SW500-Series Cabinet Installation and User’s Guide for new loading sequence illustrations for cabinets using 4 GB or larger disk drives.
Figure 3–4 SW500-Series Cabinet Loading Sequence CI BULKHEAD FRONT-TO-REAR SCSI-2 CABLE ROUTING POINTS MOUNTING LOCATIONS MOUNTING LOCATIONS HOLE #2 STORAGE POSITION S5 HOLE #8 STORAGE POSITION S4 CDU B HOLE #14 STORAGE POSITION S3 STORAGE POSITION S9 HOLE #14 HOLE #20 CONTROLLER POSITION C1 STORAGE POSITION S8 HOLE #20 HOLE #26 STORAGE POSITION S1 CONTROLLER POSITION C2 STORAGE POSITION S7 HOLE #26 STORAGE POSITION S2 STORAGE POSITION S6 HOLE #32 HOLE #32 CABINET FRONT CDU A INTER
Figure 3–5 SW500-Series Controller/Storage Cabinet Shelf and Tape Drive Locations MOUNTING LOCATIONS MOUNTING LOCATIONS HOLE #3 TAPE TAPE POSITION T2 POSITION T1 HOLE #14 STORAGE POSITION S3 HOLE #20 CONTROLLER POSITION C1 HOLE #26 HOLE #32 STORAGE POSITION S1 FRONT-TO-REAR SCSI-2 CABLE ROUTING POINTS TAPE TAPE POSITION T1 POSITION T2 CI BULKHEAD CI BULKHEAD CDU B INTERNAL CI CABLE ROUTING POINT STORAGE POSITION S2 CABINET FRONT HOLE #3 CDU A CONTROLLER POSITION C2 STORAGE POSITION S5 STOR
Figure 3–6 SW300-Series Deskside RAID Enclosure CXO-4268A-MC 3.2.4 Shelves The following configuration rules apply to the arrangement of controller and storage shelves.
3.2.4.1 BA350-Series BA350-series shelves can be arranged in any SCSI–2 legal configuration, subject to the following rules: • No more than a single extension from one BA350–SB storage shelf is permitted. The two BA350–SB shelves must be physically adjacent to each other. Intermixing 5¼-inch SBBs and 3½-inch SBB is permitted per StorageWorks configuration rules. Figure 3–7 shows an example of storage shelves in a single extension configuration.
Figure 3–8 Adjacent Devices on a Single Port BA350-MA POWER POWER HSJ40 CONTROLLER TZ867 TZ867 CXO-3751A-MC • When using a 1.0 meter cable with a controller in the lower controller shelf position (C1) in the front of the SW800 or SW500 cabinet, all front-mounted shelves can be reached. The 2.0 meter cable reaches all shelves, but does not permit shelf jumpering. (Refer to Figure 3–1.
Figure 3–9 SW300-Series Cabinet Controller/Storage Shelf POWER A POWER B POWER SUPPLY SBB DUAL-SPEED BLOWER STORAGE DEVICE SBB AC POWER ENTRY CONTROLLER A ENVIRONMENTAL MONITOR UNIT HS ARRAY CONTROLLER AC POWER ENTRY CONTROLLER B CXO-4305A-MC This shelf contains a single backplane for controller-to-storage connections. No external SCSI device cables are needed, and very little physical configuring is required as described in Section 3.4.7.
3.3.1 Nonredundant HS Array Controller Configurations The following considerations apply to nonredundant controller configurations: • A nonredundant controller must be installed in the controller slot furthest from the BA350–MA controller shelf ’s SCSI connectors. This slot is SCSI ID 7. By using SCSI ID 7, SCSI ID 6 (the other controller slot) is available as an additional ID on the storage shelf.
3.3.3 Optimal Performance Configurations For optimal performance, configure to the following guidelines: • Balance the number of devices on each port of the controller. For example, for 18 (35) SBBs attach three devices on each of the six ports. This maximizes parallel activity across the controller’s available ports. Figure 3–10 is an example of how to balance devices across ports. • Intermixing higher and lower performance devices on each port is beneficial.
Figure 3–10 Balanced Devices Within Storage Shelves UNBALANCED (6 DEVICES/PORT ON 3 PORTS) BALANCED (3 DEVICES/PORT) BA350-MA BA350-MA POWER POWER POWER POWER 6-PORT CONTROLLER BA350-SB BA350-SB POWER 1x6T POWER POWER POWER (6) 3-1/2" SBBs 2x3T POWER POWER POWER POWER (6) 3-1/2" SBBs 2x3T BA350-SB BA350-SB POWER POWER POWER POWER (6) 3-1/2" SBBs 1x6T (6) 3-1/2" SBBs BA350-SB BA350-SB (6) 3-1/2" SBBs 1x6T 6-PORT CONTROLLER (6) 3-1/2" SBBs 2x3T CXO-3698C-MC 3.3.
Figure 3–11 shows examples of optimal configurations for RAIDset members and designated spares on separate controller ports.
3.4 Typical and Recommended Configurations The following sections describe recommended device configurations for 3½-inch and 5¼-inch SBBs. Note Intermixing disk SBBs and tape SBBs on the same controller port is permitted, provided all other configuration rules in the chapter also are obeyed. Table Conventions The following describes the designations used in the following sections.
Table 3–2 3½-Inch SBB Configurations, 6-Port Controllers Number of Devices Number of BA350–SB Shelves* Configure as** Available as 3½-Inch SBBs*** Ports Used 1-2 1 (1)2x3T 5-4 1-2 3-4 2 (2)2x3T 9-8 3-4 5-18 3 (3)2x3T 13-0 5-6 19-24 4 (2)2x3T 5-0 6 5-0 6 (2)1x6T 25-30 5 (1)2x3T (4)1x6T 31-36 6 (6)1x6T 5-0 6 37-42**** 6 (6)1x7T 5-0 6 Key for Table Conventions 2x3T refers to two (split) SCSI–2 connections, separately terminated in the shelf.
Table 3–3 3½-Inch SBB Configurations, 3-Port Controllers Number of Devices Number of BA350–SB Shelves* Configure as** Available for 3½-Inch SBBs*** Ports Used 1-2 1 (1)2x3T 5-4 1-2 3-12 2 (1)2x3T 9-0 3 (1)1x6T 13-18 3 (3)1x6T 5-0 3 19-21**** 3 (3)1x7T 2-0 3 Notes 2x3T: Two (split) SCSI–2 connections, separately terminated in the shelf. The devices appear as IDs 0, 2, 4, and 1, 3, 5. 1x6T: Single path SCSI–2 connection terminated in the shelf. The devices appear as IDs 0 through 5.
• When five ports (SW800) or two ports (SW500) have doubled shelves for 5¼-inch SBBs (4+2), TZ8x7 tapes cannot be connected or even mounted in the cabinet because all or most (front) shelf locations are needed for the 5¼-inch SBBs. 3.4.4 5¼-Inch SBB Recommended Configurations Tables 3–4 and 3–5 list some recommended configurations for 5¼-inch SBBs exclusively.
Table 3–5 5¼-Inch SBB Configurations, 3-Port Controllers Number of Devices Number of BA350–SB Shelves* Configure as Available for 5¼-Inch SBBs** Ports Used 1-2 1 (1)2x3T 1-0 1-2 3-4 2 (1)2x3T 1-0 3 (1)1x6T 5-6 3 (3)1x6T 1-0 3 7-8 4 (3)1x6T 1-0 3 9-10 5 1-0 3 1-0 3 (1)1x6J (3)1x6T (2)1x6J 11-12 6† (3)1x6T (3)1x6J †Cannot be configured in SW500-series cabinets.
Table 3–6 Small Shelf Count Configurations, 6-Port Controllers Number of Devices Number of BA350–SB Shelves* Configure as Ports Used 1-6 1 1x6T** 1 7-12 2 1x6T 2 13-18 3 1x6T 3 19-24 4 1x6T 4 25-30 5 1x6T 5 31-36 6 1x6T 6 37-42*** 6 1x7T 6 Key for Table Conventions * Consult the StorageWorks Solutions Shelf and SBB User’s Guide for BA350–SB shelf information. ** T indicates that the shelf is terminated. *** Nonredundant controller and power configurations (not recommended).
Table 3–8 SW300-Series Cabinet Shelf Configuration Field Replaceable Unit Minimum Maximum StorageWorks building block (SBB) shelf power supplies 4 8 1 24 3½-Inch storage SBBs 0 5¼-Inch storage SBBs 01 8 HS array controllers 1 2 Controller cache modules 0 2 Environmental monitor units (EMUs) 1 2 AC power entry controllers 1 2 Dual speed blowers 8 8 1 Minimum of one for both 3½-inch and 5¼-inch SBBs.
Figure 3–12 SW300-Series Cabinet Shelf SCSI Buses id 3 id 3 id 3 id 3 id 3 id 3 SHELF 4 id 2 id 2 id 2 id 2 id 2 id 2 SHELF 3 id 1 id 1 id 1 id 1 id 1 id 1 SHELF 2 id 0 id 0 id 0 id 0 id 0 id 0 SHELF 1 id 6 id 6 id 6 id 6 id 6 id 6 CONTROLLER SLOT 6 id 7 id 7 id 7 id 7 id 7 id 7 CONTROLLER SLOT 7 PORT 1 PORT 2 PORT 3 PORT 4 PORT 5 PORT 6 CXO-4315A-MC 3.
3.6 CD–ROM Restrictions HSOF supports CD–ROM readers with the following restrictions: • CD–ROM devices must not be removed or inserted with power applied to the device shelf. • CD–ROMs packaged two per SBB must not be placed in a split but BA350–Sx shelf. Placing such SBBs on split buses causes adverse effects during device warm swap and bus quiesce operations. • CD–ROM devices must not be mixed with other device types (disks, tapes, and so forth) on the same controller SCSI–2 device port.
3.7.3 HSZ Array Controller Host Adapter Support The following host adapters are currently supported for HSZ array controllers: • KZTSA (for DEC 3000 systems). See the HSZ array controller release notes for restrictions. • KZMSA (for DEC 7000 and DEC 10000 systems via DWZZA) • PMAZC (for DEC 3000 system via DWZZA) Note The KZPSA and KZPBA were certified with the DEC OSF/1 Version 3.2 operating system.
4 Installation 4.1 Customer Site Preparation Site planning and preparation are necessary before installing an HS array controller subsystem. Site preparation activities should have been completed before you received your subsystem order. However, if your planning or preparation was incomplete, complete all required site preparation before you begin the installation process. WARNING To prevent damage to equipment and personnel, make sure all power sources meet the specifications required for this equipment.
4.1.1 Power and Power Cord Requirements Before installing your controller subsystem, ensure that the correct power cable (frequency and voltage) for your site is attached to the cabinet’s cable distribution unit (CDU), and that the power requirements for your country and your site have been met at the cabinet level. For specific information about power cord plugs, refer to the StorageWorks cabinet specific installation and user’s guides listed in this chapter and in the preface of this manual. 4.1.
Table 4–1 (Cont.) StorageWorks Environmental Specifications Condition Specification Maximum Nonoperating Environment (Range) Temperature Relative humidity Nonoperating Altitude –40° to +66°C (–40° to +151°F) (During transportation and associated short–term storage) 8% to 95% in original shipping container (noncondensing); otherwise, 50% (noncondensing) From –300 m (–1000 ft) to +3600 m (+12,000 ft) MS 4.
4.2.3 Electrostatic Discharge Protection Guidelines This section describes the necessary precautions and procedure for protecting the controller subsystem components against electrostatic discharge (ESD). ESD is a common problem for any electronic device and may cause lost data, system down time, or other problems. The most common source of static electricity is the movement of people in contact with carpets and clothing materials. Low humidity allows a large amount of electrostatic charge to build up.
4.3.1 Module Handling Guidelines When handling controller or cache modules, use the following ESD grounding procedure: CAUTION Use ESD grounding guidelines when handling a controller, cache module, or program card, or damage to the modules could result. Specific safety precautions must be taken when handling write-back cache modules. Therefore, only qualified service personnel can install or replace write-back cache modules. 1. Obtain and attach an ESD wrist strap to your wrist.
CAUTION The program card ESD shield must remain installed over the program card during controller operation to avoid electrostatic discharge that can cause the contents of the program card to be erased. Figure 4–1 Location of Program Card Eject Button MOUNTING SCREWS HSJ40 PROGRAM CARD EJECT BUTTON CI (HOST) CABLE MOUNTING SCREWS CXO-4118A-MC 4.3.3 Cabling Guidelines Preplanning your cabling needs and adhering to cable handling guidelines ensures proper operation of your controller subsystem.
4.3.3.1 CI Host Port Cable Handling Guidelines for HSJ Array Controllers When handling or moving CI cables with the power on, it is very important that the internal CI cable does not become grounded. This means that no metal can touch the silver plug portion of these cables, except a CI host cable connector. CAUTION DO NOT let the silver plug portion of the internal CI cable become grounded (touch metal). If this cable should become grounded, damage to the equipment can result.
Newer connectors have a notch in the connector to allow access with a very small straight-edge screwdriver. Figure 4–2 HSD30 Host Port Connector Mating Guide HSD30 HSD30 CONNECTOR MATING GUIDE CXO-4357A-MC Maximum DSSI Host Port Cable Lengths The maximum DSSI host port cable lengths are the following: • The maximum DSSI host port cable length between DSSI nodes for HSD30 array controllers is 9.14 meters (30 feet). • The overall (end-to-end) maximum configuration DSSI host port cable length is 18.
4.3.3.3 SCSI Host Port Cable Handling Guidelines for HSZ Array Controllers SCSI host port cables may be removed or replaced with power applied. However, if you must remove the SCSI host port cables for any reason while power is applied, do the following: • If it is at the end of the SCSI host bus, leave the SCSI host port cable and the terminator connected to the trilink when you remove the trilink from the controller’s front bezel.
Note Controller shelf #4 can use 2.0 meter SCSI–2 cables to the front or back shelves in an SW800-series data center cabinet. • SCSI buses are internal to the backplane in an SW300-series desktop enclosure. No device cables are required. 4.4 Unpacking Your Subsystem When delivered, your controller subsystem is packed in a carton and attached to a shipping pallet. Upon receipt of your subsystem, perform the following tasks: • Check the carton and pallet for signs of shipping damage.
Figure 4–3 shows a version of a preconfigured HS array controller subsystem in an SW500-series cabinet. The SW500-series cabinet has the capacity to hold 10 shelves (either device or controller) mounted horizontally. Six shelves can be installed in the front of the cabinet, and four shelves can be installed in the back of the cabinet.
Figure 4–4 shows a version of a preconfigured HS array controller subsystem in an SW800-series data center cabinet. An SW800-series data center cabinet has the capacity to hold up to 18 storage shelves (mounted horizontally) and four controller shelves (mounted vertically). Nine storage shelves can be installed in the front of the cabinet, and nine storage shelves can be installed in the back of the cabinet.
Figure 4–5 shows an SW300-series deskside RAID enclosure. This cabinet holds one storage/controller shelf with a single backplane for controller-to-storage connections. No external SCSI device cables are needed. The SW300-series enclosure holds up to 24 3½-inch SBBs or up to 8 5¼-inch SBBs, and up to two controller and two cache modules.
• Measure cabinet capacities by the number of individual shelves that can be installed into the cabinet (for example, BA350–SB storage shelves). • Abide by the cabinet-specific installation sequence to maintain gravitational stability. Failure to follow the sequence may result in tipping over the cabinet. • Shelf configuration rules apply to all cabinets unless otherwise noted.
4. Plug the cabinet power cord into the proper wall outlet for your subsystem’s power requirements (for example, 60 Hz 220 volts, 50 Hz 240 volts). 5. Plug a maintenance terminal’s EIA cable into the maintenance terminal port on the front bezel of your controller (for defining your subsystem’s initial configuration parameters). See Section 4.6.1 for maintenance terminal setup information. 6. Turn on the power to your maintenance terminal.
For HSD30 array controllers—You should have connected the appropriate host port cables and/or terminators to the supplied trilink connector on the front bezel of your controller BEFORE you applied power. Figure 4–6 shows the trilink connectors, DSSI host port cable, DSSI host port cable jumper, and terminator for a dual-redundant HSD30 array controller configuration. See Section 4.6.2.2 for detailed instructions for installing these components.
For HSZ array controllers—Connect the appropriate host port cables and terminators to the trilink connector block on the front bezel host port connectors of your controller. Figure 4–7 shows the connection of a SCSI host port cable and terminator to the trilink connector block on a nonredundant HSZ array controller. Figure 4–8 shows the connections for a dual-redundant configuration.
Figure 4–8 SCSI Host Port Cable Connection—Dual-Redundant Configuration HSZ40 HSZ40 TRILINK CONNECTOR TERMINATOR SCSI HOST CABLE CXO-4359A-MC Refer to Sections 4.6.2.1, 4.6.2.3, and 4.6.2.2 for specific information for connecting your controller’s host port cables. Note Do not connect the controller end of any internal host port cables to your controller unless the controller’s initial parameters have been set and the host ports have been enabled.
Figure 4–9 Internal CI Cable with External CI Cables Attached for HSJ Array Controllers REVERSE VIEW (TO FRONT OF HSJ CONTROLLER) INTERNAL CI CABLE CHANNEL # CLUSTER A NODE PATH TRANSMIT RECEIVE (RX) (TX) Cl ter GREEN LABEL r us TX Node Clus te No de th Path Pa RX RED LABEL LEGEND B EXTERNAL CI CABLES CXO-3753B-MC 4.6.1 Connecting a Terminal to the Maintenance Terminal Port You do not need a locally connected maintenance terminal for normal operations.
3. Connect the other end of the terminal cable to the MMJ terminal port on the controller’s front bezel. Note For HSZ40–Bx array controllers (with offset MMJ connectors), install the short phone cable (Digital part number 17–03511–04) and the center-tooffset coupler (Digital part number 12–43346–01). Plug the terminal cable into the coupler. 4. Turn the terminal power switch to the ON position. 5. Set the terminal at 9600 baud, with 8 data bits, 1 stop bit, and no parity.
4.6.2.1 Installing Host Port Cables for HSJ Array Controllers Preconfigured HSJ array controller subsystems come with the internal host port cable preinstalled. When adding a second controller to a nonredundant (single) controller configuration, do the following: • Position and route the new internal CI host port cable using the same path as the host port cable for the first controller.
• To turn on the CI host port paths for a dual-redundant controller configuration, enter the following CLI commands: CLI> SET THIS_CONTROLLER PATH_A CLI> SET THIS_CONTROLLER PATH_B CLI> SET OTHER_CONTROLLER PATH_A CLI> SET OTHER_CONTROLLER PATH_B If you have a single controller, use only the SET THIS commands. 4.6.2.
5. To turn on the DSSI host port paths, enter the following CLI commands (or command, if you are merely enabling one controller): CLI> SET THIS_CONTROLLER PATH CLI> SET OTHER_CONTROLLER PATH 4.6.2.3 Installing Host Port Cables, Trilinks, and Terminators for HSZ Array Controllers The following procedure describes how to connect host port cables, trilink connector blocks, and terminators for HSZ array controllers: 1.
• Route the other end of the SCSI–P cable to the host. If you use a desktop DWZZA–AA signal converter, do the following: • Plug the SCSI host port cable coming from the controller’s trilink into the connector on the front of the DWZZA–AA signal converter. • Plug another SCSI host port cable into the back of the DWZZA and route the cable to the host. 5. For 16-bit FWD adapters, route the host port cable directly from the controller’s trilink connector, through the cabinet, and to the host port adapter.
The second controller will occupy the SCSI ID #6 slot. The second controller will not function correctly if there are attached devices in slot 6 of any storage shelf for any port. Check that each port has no more than six SCSI–2 devices at ID numbers 0 through 5. Enter the SHOW DEVICES command and look at the Port column for each device to determine the number of devices configured on each port. Figure 4–11 is an example of a dual-redundant HSJ40 configuration with controllers in SCSI ID slots 6 and 7.
Figure 4–11 Controller ID Numbers PRIMARY POWER SUPPLY SECONDARY POWER SUPPLY BA350-MA SHELF HSJ40 CONTROLLER ID6 HSJ40 CONTROLLER ID7 CXO-3662A_PH CAUTION In a dual-redundant configuration with cache modules, both cache modules must have the same number of megabytes and both must be the same cache type (both read or both write-back). The controller firmware version must be the same for both controllers for proper operation of the subsystem.
When the new controller and cache modules are installed, set your initial configuration parameters for the new controller using the procedure in Section 5.2.5. Do not install your host port cables until the initial parameters are set. 4.8 Installing the Program Card This section describes how to insert or remove the program card. Figure 4–12 shows the location of the program card and its associated eject button.
Figure 4–12 Program Card and Its Eject Button (ESD Shield) PROGRAM CARD HSJ40 PROGRAM CARD EJECT BUTTON CXO-4333A-MC Use the following procedures to replace your program cards in the event that you wish to load new firmware or if you had to remove the program card for some reason: Program Card Upgrade—Nonredundant Configuration An HS operating firmware upgrade causes a brief interruption in service to attached drives.
6. While holding the OCP reset (//) button, insert the new program card, pressing the card in until the eject button extends outward almost even with the program card, then release the reset (//) button. The controller restarts. Note Following the controller restart, approximately 60 seconds pass before the controller accepts CLI commands. This delay does not affect unit availability to the host.
6. Replace the ESD shields over the program cards by pushing the two plastic push pins on each side of the two shields. 7. Remove the ESD grounding strap. 4.9 Upgrading Your HS Array Controller Subsystem Components The following is a list of components that you can upgrade in your HS array controller subsystem: • Program card—Upgrade your firmware by replacing the program card with a program card containing the new firmware version.
How to Determine Your Cache Module Type To determine your cache module type (version type and number of megabytes), enter the SHOW THIS_CONTROLLER or SHOW OTHER_CONTROLLER command at the CLI> prompt as appropriate. Note The original HSJ40 array controllers shipped with Version 1 read cache modules. The HSJ30, HSD30, and HSZ array controllers all ship with Version 2 read cache modules. You must have HS operating firmware Version 1.4 or greater to operate with a Version 2 read cache module.
Example 4–1 shows an HSJ40 array controller subsystem without a cache module: Example 4–1 No Cache Module Installed MASS>SHOW THIS_CONTROLLER Controller: HSJ40 (C) DEC ZG303FF082 Firmware v25J, Hardware 0000 Not configured for dual-redundancy SCSI address 7 Time: 14-FEB-1995 15:21:10 Host port: Node name: MASS, valid CI node 4, 32 max nodes System ID 420010042D4D Path A is ON Path B is ON MSCP allocation class 6 TMSCP allocation class 6 CI_ARBITRATION = ASYNCHRONOUS Cache: No cache MASS> Example 4–2 shows a
Example 4–3 shows a controller subsystem with a Version 2, 32-MB write-back cache module: Example 4–3 Version 2 32-MB Write-Back Cache Module KATHYM>SHOW THIS_CONTROLLER Controller: HSJ40 (C) DEC ZG313FF106 Firmware V25J, Hardware 0000 Configured for dual-redundancy with ZG31900458 In dual-redundant configuration SCSI address 7 Time: 14-FEB-1995 15:21:10 Host port: Node name: HSJA1, valid CI node 3, 32 max nodes System ID 420010031B85 Path A is ON Path B is ON MSCP allocation class 4 TMSCP allocation class
Table 4–3 Read and Write-Back Cache Upgrade Ordering Information for HSJ40 Array Controllers If you have...1 And you want... Order...
Table 4–6 Read and Write-Back Cache Upgrade Ordering Information for HSZ40–Ax and HSZ40–Bx Array Controllers If you have... And you want... Order...
5 Controller Operations 5.
Figure 5–1 details the initialization process for both single and dual-redundant controller configurations.
2. The firmware executive (EXEC) initializes and a call is made to the diagnostic and execution monitor (DAEMON) to complete the subsystem self-test diagnostics. When executing these diagnostics, DAEMON uses the services provided by the EXEC to handle any interrupts and errors generated during the tests. Errors detected during these tests are displayed on the OCP LEDs.
3. During the initialization, the controller interrogates the entities that it has established connections with, including other controllers in the subsystem. The controllers can perform activities with the entities, such as illuminating the device activity LED (green) on a disk drive’s carrier while the drive is spinning up or down. 4. When the initialization sequence on all entities completes, the controller begins data transfer and other operations with the hosts. 5.1.
Note The CLI> prompt may be factory-set to reflect your array controller’s model, such as, HSJ>, HSD>, or HSZ>. Use either the SET THIS_CONTROLLER or SET OTHER_CONTROLLER command with the PROMPT= qualifier if you want to set your prompt name (see Appendix B) to something other than the factory setting. 5.2.2 CLI Command Sets CLI consists of the following six command sets: 1. CLI failover commands—These commands are used to support dualredundant controller configurations. 2.
exercisers. There are additional commands used to run local programs such as, CFMENU, FLS, and so forth. Remember these two rules when using the CLI: • Not all configuration parameters need to be specified on one line. They can be entered by using multiple SET commands. • Only enough of each command needs to be entered to make the command unique (usually three characters). For example, SHO is equivalent to SHOW. 5.2.3 How to Exit CLI If you are using a maintenance terminal, you cannot exit the CLI.
1. Set the MAX_NODES parameter (HSJ controllers only): CLI> SET THIS_CONTROLLER MAX_NODES=n where n is 8, 16, or 32. 2. Set a valid ID for the controller (all controller models): CLI> SET THIS_CONTROLLER ID=n Where n has the following values: • The HSJ controller’s CI node number (0 through (MAXNODES-1)). CAUTION If the CI node number, but not the SCS node name, is changed, a complete VMScluster system reinitialization must be performed before the controller and its attached devices will be recognized.
6. Restart the controller by pressing the OCP reset (//) button on the controller’s front bezel, or enter the following command at the CLI> prompt to restart the controller (for all controller models:) CLI> RESTART THIS_CONTROLLER 7. Enter the SHOW THIS_CONTROLLER command to determine whether the preceding parameters are set correctly (for all controller models:) CLI> SHOW THIS_CONTROLLER FULL 8.
connecting or disconnecting the DSSI connector at all times, even if you have the mating guide installed. Defining Devices and Storagesets The following steps configure devices into storagesets and storagesets into logical units seen by the host. If these steps are not completed, the host cannot access any devices. These steps can be done from a virtual (host) terminal, or from the same maintenance terminal that you used to set initial parameters.
2. Add storagesets by using commands similar to the following: CLI> ADD STRIPESET container-name container-name1 ... CLI> ADD RAIDSET container-name container-name1 ... CLI> ADD MIRRORSET container-name container-name1 ... For example: CLI> ADD STRIPESET STRIPE0 DISK0 DISK1 DISK2 DISK3 CLI> ADD RAIDSET RAID99 DISK4 DISK5 DISK6 CLI> ADD MIRRORSET MIRR1 DISK7 DISK8 DISK9 Refer to Chapter 6 and Appendix B for examples for adding storagesets.
5.2.5 Setting Configuration Parameters for a Dual-Redundant Controller Configuration If you have a new dual-redundant controller configuration, perform the following steps at initial installation to set up your controller’s configuration parameters.
where xxxxxx is a one- to six-character alphanumeric name for this node that must be enclosed in quotation marks and begin with an alphabetic character. Note Each HSJ or HSD30 controller’s SCS node name must be unique within its VMScluster system. 4. Set the MSCP allocation class on both controllers (HSJ and HSD30 controllers): Note Provide different (unique) allocation class values for every pair of dualredundant controllers in the cluster.
7. (All controller models:) Restart both controllers by pressing the OCP reset (//) button on each controller’s front bezel, or enter the following commands at the CLI> prompt in the sequence given: CLI> RESTART OTHER_CONTROLLER CLI> RESTART THIS_CONTROLLER 8. Enter the following commands at the CLI> prompt to determine whether the preceding parameters are set correctly (all controller models): CLI> SHOW THIS_CONTROLLER FULL CLI> SHOW OTHER_CONTROLLER FULL 9.
11. HSZ Array Controllers: Set preferred paths to balance the load and improve the performance of your HSZ array controller as follows: After setting the total number of SCSI target IDs for the pair, set the preferred IDs to define which targets will be assigned to each controller.
cables attached to warm swap your HSD30 array controller. You cannot perform controller warm swap if you have this situation. Defining Devices and Storagesets To automatically configure devices on the controller, use either the Configure (CONFIG) utility or CFMENU utility described in Chapter 7 (for all controller models). For manual configuration, the following steps add devices, storagesets, and logical units. If these steps are not completed, the host cannot access the devices.
Refer to Chapter 6 and Appendix B for examples for adding storagesets. If you do not need storagesets for your configuration, skip this step. Note If you add a removable media device to an HSJ or HSD30 controller, it is not known to the host until one of the following occurs: the media is loaded into the device, the host is reinitialized, or the virtual circuit is broken and reestablished. Remember that CD–ROMs are always set TRANSPORTABLE and are seen by the OpenVMS operating system as DKA devices. 3.
5.3 Using the TRANSPORTABLE and NOTRANSPORTABLE Qualifiers Before initializing a device using the CLI INITIALIZE command, make sure you know the status of that device. Ask yourself the following questions before proceeding with a device initialization: • Will this device be used on HSx and HSC controllers, or not moved at all? • Will this device be used on an HS array controller or HSC controller? (If so, set it NOTRANSPORTABLE.) • Did it come from a non-HS or non-HSC k.
• Run DILX—Use the default answers to the test questions (see Chapter 7). This tests all disk devices in your subsystem. DILX tests logical units that can consist of storagesets or physical devices. Error reports identify the logical units, not the physical devices. Therefore, if errors occur while running against a storageset, the storageset should be reconfigured as individual devices, and then DILX run against the individual devices.
Figure 5–2 HSJ40 Operator Control Panel Reset and Port Buttons PROGRAM CARD RESET BUTTON (GREEN LED) // 1 2 3 4 PORT QUIESCE BUTTONS (AMBER LEDs) PROGRAM CARD EJECT BUTTON 5 6 CXO-3664A_PH Controller Operations 5–19
Figure 5–3 HS Array Controllers Solid OCP LED Error Codes Reset Off 1 2 3 4 5 6 Description of Error Action 3F DAEMON hard error. Replace controller module. 3E Repeated firmware bugcheck. Replace controller module. 3D NVMEM version mismatch. Replace program card with later version. 3C NVMEM write error. Replace controller module. 3B NVMEM read error. Replace controller module. 3A NMI error within firmware bugcheck. RESET (//) the controller.
Figure 5–4 HS Array Controllers Flashing OCP LED Error Codes Reset Off 1 2 3 4 5 Lit continuously 6 Description of Error Action 01 Program card EDC error. Replace program card. 04 Timer zero in the timer chip will run when disabled. Replace controller module. 05 Timer zero in the timer chip decrements incorrectly. Replace controller module. 06 Timer zero in the timer chip did not interrupt the processor when requested. Replace controller module.
Figure 5–4 (Cont.) HS Array Controllers Flashing OCP LED Error Codes Reset Off 1 2 3 4 5 Lit continuously 6 Action Description of Error 13 The controller DRAB chip failed to detect forced parity, or detected parity when not forced. Replace controller module. 14 The controller DRAB chip failed to verify the EDC correctly. Replace controller module. 15 The controller DRAB chip failed to report forced ECC. Replace controller module.
Figure 5–4 (Cont.) HS Array Controllers Flashing OCP LED Error Codes Reset Off 1 2 3 4 5 Lit continuously 6 Action Description of Error 21 The required amount of memory available in the pool area is insufficient for the controller to run. Replace controller module. 23 The required amount of memory available in the buffer area is insufficient for the controller to run. Replace controller module.
indicated FRU. If the error indication changes, look up that code in Figure 5–3 or Figure 5–4 and replace the indicated FRU. Note The OCP LEDs flash once per second to indicate normal operations, remain on solid or flash three times per second to indicate an error condition. After initialization is complete, the port buttons and associated LEDs are used as Bus Quiesce Request buttons in device warm swap removal and replacement procedures. 5.
• Turns on the fault (amber) LED on the front upper right corner of the SW300 cabinet • Activates an audible alarm • Increases the speed of the blowers Figure 5–5 shows the EMU front panel LEDS which display information when the subsystem is turned on or encounters a problem. Figure 5–5 EMU Front Panel LEDs and Switches BLOWERS AUDIBLE ALARM SWITCH 1 5 SYSTEM OK 2 6 3 4 7 8 TEMPERATURE SHELF FAULT BLOWERS CXO-4310A-MC 5.8.
Table 5–1 lists the EMU front panel buttons and LEDs, their functions and error descriptions. Table 5–1 EMU Front Panel LEDS/Switches and Error Descriptions When the ... Is The System ... System OK (green) LED ON Is in the normal operating state. Shelf Fault (amber) LED OFF Blower 1 through 8 (amber) LEDs OFF Temperature (amber) LEDS OFF Audible Alarm OFF Audible Alarm ON Has a failed FRU. Check the EMU control panel LEDS to determine which FRU has failed.
5.9 Description of Device Warm Swap Device warm swap is a sequence of quick removal and insertion steps, allowing an operator to safely remove a device and insert another device in its place, or to add a new device. The reason for performing the following steps is to protect data integrity for other shelf devices, and to reduce the chance of an operator causing a port to be unusable for a long period of time (which could render several devices inaccessible). Only one port may be quiesced at any time.
RAIDsets that are not already running as reduced will automatically adjust to the removal of a device (the RAIDset will go reduced). In this case, there is no need to dismount the RAIDset unit. However, you must dismount the RAIDset unit if the RAIDset is already reduced. Mirrorsets that have at least one NORMAL or NORMALIZING member other than the one being removed automatically adjust to removed of the device. In this case, there is no need to dismount the affected unit.
Figure 5–6 SBB Warm Swap CXO-3611B-PH While the OCP LEDs are flashing their alternating pattern, the shelf SBB LEDs for all disk SBBs on that port also flash. (Although tape drives do not always have LEDs, remove them at this time if you are using warm swap to remove a tape drive. Some tape drives cannot be warm swapped.) The interval between the quiesce request and the actual halting of I/O can vary from zero seconds to 1 minute, depending on the load, device type, and cache status.
After the SBB is inserted, its port LED flashes for approximately 15 seconds and then turns off, unless other configuration errors exist on that port. Normal operation resumes on the port. If a new device is added in a previously unused slot, that port’s LED remains on until the device is added to the controller’s configuration. If a tape device is placed in a slot where a disk device was previously installed, the port LED remains on until the tape device is added to the controller’s configuration.
Note HSD30 and HSZ controllers: You cannot effectively warm swap some earlier HSD30 or HSZ array controllers (in slot SCSI ID 7 in BA350–MA shelves, SCSI ID 6 in SW300-series cabinet shelves), because interference from the companion controller’s trilink connector prevents removal and replacement.
The C_SWAP utility can also be used to add a second controller and cache module to a nonredundant controller configuration, or for upgrading cache modules. The C_SWAP program first goes out and looks at the hardware configuration in the controller shelf. If it sees a dual-redundant configuration it assumes a controller or cache module is going to be removed. If it sees a nonredundant configuration it assumes a controller and cache module are going to be added.
5. Enter the following command: HSx> RUN C_SWAP The system responds with: Controller Warm Swap, Software Version -V2.0 *** Sequence to REMOVE other HSJ40 has begun. *** Do you wish to REMOVE the other HSJ40 Y/N [N]? YES 6. Enter ‘‘Y’’ to continue the procedure. Will its cache module also be removed Y/N [N]? YES 7. Enter ‘‘Y’’ only if you are removing the controller’s cache module as well.
The replacement configuration must contain both a cache module and an HS array controller module. The replacement HS array controller module must have its program card removed during the insertion process. This eliminates the possibility of prematurely starting the C_SWAP program before the module is completely installed. Once the second controller is detected, C_ SWAP displays a restart procedure and exits. Insert the program card into the newly replaced controller module to complete the restart procedure.
1) Enter the RESTART OTHER_CONTROLLER command. 2) Press and hold in the Reset (//) button while inserting the program card. 3) Release Reset (//); the controller will initialize. 4) Configure new controller by referring to the StorageWorks Controller User’s Guide.
Table 5–2 Controller Module Removal Step Description 1 Unlock and open the cabinet doors (SW800-series) using a 5/32-inch Allen wrench. 2 Ground yourself to the cabinet grounding stud (refer to Figure 5–7). 3 Unsnap and remove the program card ESD shield that covers the program card on the failed controller module. 4 Remove the program card by pushing the eject button next to the card. Pull the card out and save it for use in the replacement controller module.
Figure 5–7 Cabinet Grounding Stud SW800-SERIES CABINET GROUNDING STUD CXO-4116A-MC Controller Operations 5–37
Figure 5–8 Four Front Bezel Screws, HSJxx Controller MOUNTING SCREWS HSJ40 PROGRAM CARD EJECT BUTTON CI (HOST) CABLE MOUNTING SCREWS CXO-4118A-MC 5.13 How to Physically Remove a Cache Module Most controller modules will have a cache module installed behind them in the controller shelf. Use the C_SWAP utility to remove, replace, or upgrade your cache module. CAUTION For safety reasons, only qualified personnel may service the write-back cache module or its batteries.
Use the following procedure to remove your read cache module: Table 5–3 Cache Module Removal Step Description 1 Ground yourself to the cabinet grounding stud (Figure 5–7) before servicing the cache module. 2 The controller module is seated in front of the cache module. Anytime you service a cache module, you must consider when it is appropriate to remove the controller module (based on considerations of configuration, down time, and so on). 3 To access the cache module, remove its controller module.
Use the procedure in Table 5–4 to replace your controller module during the controller warm swap replacement. Note Be sure to have the initial parameters that were set for the failed controller available. You will need these parameters to set the initial parameters for the new controller module. Table 5–4 Controller Module Replacement Step Description 1 Ground yourself to the cabinet grounding stud (refer to Figure 5–7). 2 You should replace the cache module now, if you removed it. Refer to Section 5.
Table 5–4 (Cont.) Controller Module Replacement Step Description 13 Snap the program card ESD shield into place by pressing the two plastic push pins on each side of the shield. 14 Remove the ESD grounding strap from the cabinet and your wrist. 15 Enter the SET FAILOVER COPY=OTHER_CONTROLLER command to copy the configuration information from the good (other) controller to this new controller. 16 Enter the SHOW THIS_CONTROLLER FULL to ensure that the dualredundant configuration has been created.
Figure 5–10 Controller Shelf Rails CACHE MODULE CONTROLLER MODULE CXO-4120A-MC 5–42 Controller Operations
5.15 How to Physically Replace a Cache Module Use the procedure in Table 5–5 to replace your cache module. CAUTION For safety reasons, only qualified personnel may remove or replace write-back cache modules. For complete instructions for replacing the write-back cache module, refer to the StorageWorks Array Controllers HS Family of Array Controllers Service Manual. Table 5–5 Cache Module Replacement Step Description 1 The controller module is seated in front of the cache module.
CAUTION If the correct steps for turning off the power to an HS array controller configuration with write-back cache are not followed, there is a potential for loss of data that may exist on any devices connected to the HS array controllers. 1. Use the proper procedures for shutting down your operating system if the host system is also going to be turned off.
After restoring power to the system, and restarting the controllers, check the battery status before accessing any write-back enabled units from the operating system. The battery status must say ‘‘good,’’ unless the CACHE_POLICY=B qualifier is set. Remember that if you have the CACHE_ POLICY=B set, and you decide to use your RAIDsets or mirrorsets, you risk losing data if power should be lost before the batteries have recharged.
6 Working with RAID Arrays 6.1 HS Array Controller Family RAID Overview Note For discussions in this manual, the term stripesets refers to RAID level 0, the term mirrorsets refers to RAID level 1, and the term RAIDsets refers to RAID level 5 (with RAID level 3 features).
6.1.1 RAID Level 0 RAID level 0 is known as striping. Striping spreads data across multiple disks, breaking the user data into segments designated as ‘‘chunks.’’ In a four disk stripeset, A, B, C, and D, for example, the first chunk is written on disk A, the second on disk B, the third on disk C, the fourth on disk D, the fifth on disk A, and so on. CAUTION If any member of a RAID level 0 stripeset fails, all data is lost from the entire set.
• Flexible policy options for determining both how read requests are satisfied and the speed of copying when a new member is being added. In addition the mirroring option supports a feature which allows a system administrator to create an identical copy of any HS array controller disk or stripeset unit, and then dissociate it from the original. This feature is called ‘‘cloning.
Note Chunk size is set with the CLI INITIALIZE CHUNKSIZE= command (refer to Appendix B). 6.1.5 RAID Level 5 RAID level 5 stripes data and rotates parity across all disks in the RAIDset. The controller combines incoming data with existing parity data. RAID level 5 is suited for applications whose I/O loads consist predominantly of a large number of asynchronous read requests. Transaction processing and office automation applications often fall into this category.
• Nominal membership—The desired number of mirrorset members when the mirrorset is fully populated with active devices. If a member is removed from a mirrorset, the actual number of members may fall below the ‘‘nominal’’ target. • NORMAL member—A mirrorset member whose entire contents is guaranteed to be the same as all other NORMAL members. All NORMAL members are exactly equivalent.
When a RAIDset is initialized using the INITIALIZE command, the controller does not take the lengthy period of time to make all the parity blocks consistent with the data. Instead, the controller marks all the parity blocks as bad and starts a reconstruct. The reconstruct recalculates and rewrites the parity blocks and marks them as good. This process allows the RAIDset to be used immediately. All new data written to the RAIDset is immediately fully redundant.
6.3 RAIDset and Mirrorset Rules and Important Information The following list gives rules to remember about RAIDsets and mirrorsets: • You must always have a write-back cache module when creating RAID level 5 RAIDsets or mirrorsets. • Do not attempt to use any RAIDset or mirrorset commands with mismatched cache modules. Both cache modules must be write-back cache, and both must have the same number of megabytes.
• Your RAIDsets can achieve the high performance characteristics of RAID level 3 provided you set your RAIDset chunksize to the minimum value (16) and your application calls for large sequential I/O operations. • RAIDsets and mirrorsets can contain disks of different sizes, but the disk space used by the storageset is limited to the maximum size of the smallest disk in the RAIDset. • Place RAIDset and mirrorset members on different ports.
• If there is inactivity across the bus of the destination device for more than the time set by the CACHE_FLUSH_TIMER= qualifier (or the 10 second default), any outstanding write data is automatically flushed to the inactive devices. • Write-back cache modules must not be moved from their backplane slots unless all unwritten write cached data has been flushed. To determine whether all cache data has been flushed, use the SHOW THIS_CONTROLLER (or SHOW OTHER_CONTROLLER) command to check cache status.
• RAIDset configurations for availability and performance • RAIDset hardware requirement 6.5.1 Creating a RAIDset Enter the following commands to create a RAIDset: 1. Use the ADD DISK container-name SCSI-location command to add new disk drives to your configuration and name them. CLI> ADD DISK DISK0 1 0 0 2. Use the ADD RAIDSET container-name container-name1 container-name2 [container-nameN] command and set the appropriate replacement policy and reconstruct qualifiers.
4. Enter the ADD UNIT unit-number container-name command to create a host accessible logical unit from the RAIDset, followed by the appropriate qualifier for cache transfer sizes, preferred path, cache access, write protection, and so forth (as described in Appendix B). CLI> ADD UNIT D170 R3 Where: D170 is the unit name. R3 is the RAIDset name. 5. Enter the SHOW RAIDSETS command to display all of the RAIDsets known to the controller.
HSJ> SHOW RAIDSETS FULL Name Storageset Uses Used by -----------------------------------------------------------------------------RAID0 raidset DISK100 DISK300 DISK400 DISK500 D0 Switches: POLICY (for replacement) = BEST_PERFORMANCE RECONSTRUCT (priority) = NORMAL CHUNKSIZE = 128 blocks State: RECONSTRUCT 3% complete on member DISK500 DISK100 (member 0) is RECONSTRUCTING DISK300 (member 1) is RECONSTRUCTING DISK400 (member 2) is RECONSTRUCTING DISK500 (member 3) is RECONSTRUCTING Size: 2050353 blocks En
Enter the following CLI command to display all information about all storagesets known to the controller: HSJ> SHOW STORAGESETS FULL Name Storageset Uses Used by -----------------------------------------------------------------------------STRIPE0 stripeset DISK110 DISK210 DISK310 D1 DISK100 DISK300 DISK400 D401 Switches: CHUNKSIZE = 128 blocks State: NORMAL DISK110 (member 0) is NORMAL DISK210 (member 1) is NORMAL DISK310 (member 2) is NORMAL Size: 2050353 RAID0 raidset Switches: POLICY (for replace
Enter the following CLI commands to remove one or more disks from the spareset: HSJ> DELETE SPARESET disk-container-name0 [disk-containter-nameN] Example: HSJ> DELETE SPARESET DISK100 HSJ> DELETE SPARESET DISK100 DISK200 DISK300 Enter the following command to show the spareset: HSJ> SHOW SPARESET Name Storageset Uses Used by -----------------------------------------------------------------------------SPARESET spareset DISK310 DISK600 6.5.
To change the speed at which a RAIDset will be reconstructed when a new member is added to the RAIDset, or immediately after the RAIDset is initialized, enter one of the following commands: CLI> SET RAIDset-container-name RECONSTRUCT=NORMAL CLI> SET RAIDset-container-name RECONSTRUCT=FAST If you need to remove a disk member from a RAIDset, enter the following command: CLI> SET RAIDset-container-name REMOVE=disk-container-name For example: CLI> SET R3 REMOVE=DISK100 If the RAIDset is already in a reduced sta
6.5.7 Moving a RAIDset You may physically relocate some or all of a RAIDset’s member devices according to the following procedure: CAUTION If you lose track of the RAIDset members at any point during this procedure, you will have to attempt to restore the RAIDset by guessing where its members are installed. There is currently no way to retrace your steps using the controller or HS operating firmware. To move a RAIDset you must do the following: 1. Make note of all devices comprising the RAIDset.
If you move a RAIDset from one controller to another and you damage one member, you must specify all of that RAIDset’s members when you re-add the RAIDset to the new controller. The controller will automatically reduce the RAIDset when it discovers that one member is inoperative. Using the REDUCED Qualifier with the ADD RAIDset Command Only use the REDUCED qualifier (with the ADD RAIDSET command) when you want to move a RAIDset that is already reduced.
6.7 Planning Your Mirrorsets The following items should be considered before creating your mirrorsets: • Mirrorset size (1 to 6 members) • Mirrorset replacement policy • Mirrorset spares • Mirrorset hardware requirements—write-back cache module • Mirrorset firmware requirements—HSOF Version 2.5 firmware 6.8 Using Mirrorsets to Obtain Snapshot Copies of Data A mirrorset is a storageset consisting of multiple devices, each containing an identical copy of the same data.
2. The REDUCE command allows multiple disks to be simultaneously removed from multiple mirrorsets as a single, synchronized operation. This permits snapshot copies of stripesets and striped mirrorsets in a way that retains the integrity of the entire stripeset. 3. The NODESTROY option on the INITIALIZE command permits the snapshot copy to be configured and added as a unit (distinct from the original) without destroying any of the data. 6.
4. The extraction of the snapshot member is done with the REDUCE command. The copy must be done before the REDUCE will be honored. If multiple snapshot members are involved (as they will be if you are copying the data on a stripeset or striped mirrorset), then all copies on all members must be complete. 5. Snapshot disks must be the same size or larger than the disk members they will be used to copy. 6.
7. For each mirrorset in the unit, use the SET MEMBERSHIP command to increase the nominal membership of the mirrorset by one. This will provide the member slot for the snapshot member. CLI> SET mirrorset-container-name MEMBERSHIP=n 8. For each mirrorset in the unit, use the SET REPLACE= command to add the target snapshot disk you desire into that mirrorset. At this point, the copy to the snapshot disk will begin. CLI> SET mirrorset-container-name REPLACE=disk-device-name 9.
If the snapshot is of a stripeset, create a stripeset consisting of each of the constituent snapshot disks: CLI> ADD STRIPESET stripeset-name snapshot-disk-name-1 snapshot-disk-name-2 If the snapshot is of a simple disk or a simple mirrorset, the snapshot disk is used as a simple disk device. 2. Initialize the snapshot stripeset or device using the NODESTROY option. Only initialize the top storageset on which you are going to add the unit.
6.11.
If there are no spareset members that are at least the size of the mirrorset, then the mirrorset is left with the reduced number of members. • NOPOLICY—Allows you to turn off the autosparing capability. Allows you to remove a failing device from a mirrorset without selecting a replacement. This causes the mirrorset to run in a reduced state until a BEST_FIT or BEST_PERFORMANCE policy is selected, or a member is manually replaced in the mirrorset.
6.11.5 MIRROR disk-device-name container-name Command Allows you to convert a physical device to a one member mirrorset. CLI> MIRROR disk-device-name mirrorset-container-name For example: CLI> MIRROR DISK100 MR1 6.11.6 UNMIRROR disk_device-name Allows you to convert a one member mirrorset back to a physical device. CLI> UNMIRROR disk-device-name Refer to Section 7.3.12 for mirrorset examples performed by the CLONE utility. 6.
7 Diagnostics and Utilities This chapter contains descriptions and instructions for running the following HS array controller local programs and for using command disks: DILX TILX VTDPY FLS CONFIGURE (CONFIG) CFMENU utility Code Patch/Code Patch (CLCP) utility CLONE utility Command disks C_SWAP utility and FMU are not described in this chapter. Refer to Section 5.11 for information regarding the Controller Warm Swap (C_SWAP) utility.
7.2 Connecting to the Controller You can connect to the controller using a maintenance terminal connected to the EIA–423 terminal port, or a virtual (host) terminal using a diagnostic utility protocol (DUP) connection, or for HSZ40 array controllers, an HSZterm connection to the terminal port. 7.
Error reports identify logical units, and also may identify physical devices. Therefore, if errors occur while running against a unit, its storageset should be reconfigured as individual devices, and then DILX run again, against the individual devices. There are no limitations on the number of units DILX may test at one time. However, Digital recommends only using DILX when no host activity is present.
7.3.1.2 Interrupting DILX Execution Use the following guidelines to interrupt DILX execution: Note The circumflex symbol (^) is equivalent to the Ctrl key. You must press and hold the Ctrl key and type the character key given. Note Do not enter Ctrl/G from a VCS, because it causes VCS to terminate. VCS acts on the sequence, and the sequence never gets sent to DILX. Use Ctrl/T when interrupting DILX from a VCS. • Ctrl/G causes DILX to produce a performance summary.
Basic Function Test The basic function test for DILX executes in four phases: • Initial Write Pass—Is the only optional phase and is always executed first (if selected). The initial write pass writes the selected data patterns to the entire specified data space or until the DILX execution time limit has been reached. Once the initial write pass has completed, it is not reexecuted no matter how long the DILX execution time is set. The other phases are reexecuted on a 10-minute cycle.
User-Defined Test When you select the User-Defined test, DILX prompts you for input to define a specific test. In the User-Defined test, a total of 20 or fewer I/O commands can be defined. Once all of the commands are issued, DILX issues the commands again in the same sequence. This is repeated until the selected time limit is reached. As you build the test, DILX collects the following information from you for each command: • The I/O command name (write, read, access, erase, or quit).
Example 7–1 (Cont.) Using All Functions ENGHRN FORCE WODWND CYMBAL LUTE MASS2 MASS1 VMS HSC VMS VMS VMS HSJ HSJ V5.5 V700 V5.5 V5.5 V5.5 V20J V20J MEMBER MEMBER MEMBER MEMBER (Entered a Ctrl/C here.) DUP>SET HOST/DUP/SERVER=MSCP$DUP MASS1/TASK=DILX %HSCPAD-I-LOCPROGEXE, Local program executing - type ^\ to exit Disk Inline Exerciser - version 2.0 The Auto-Configure option will automatically select, for testing, half or all of the disk units configured.
Example 7–1 (Cont.) Using All Functions Write enable disk unit(s) to be tested (y/n) [n] ?y The write percentage will be set automatically. Enter read percentage for Random IO and Data Intensive phase (0:100) [67] ? Enter data pattern number 0=ALL, 19=USER_DEFINED, (0:19) [0] ? Perform initial write (y/n) [n] ?y The erase percentage will be set automatically.
7.3.2.2 Using All Defaults (Read-Only) Note The following example can be used for DILX runs on HSJ and HSD30 array controllers only. In Example 7–2, DILX is run using all defaults. This is executed in read-only mode. No data on the units under test will be destroyed. The entire useravailable LBN range on each disk is accessible for DILX testing. DILX was invoked from a maintenance terminal.
Example 7–2 (Cont.) Using All Defaults (Read-Only) DILX Summary at 13-JAN-1994 04:49:14 Test minutes remaining: 9, expired: 1 Unit 10 Total IO Requests 4530 No errors detected Unit 12 Total IO Requests 2930 No errors detected Reuse Parameters (stop, continue, restart, change_unit) [stop] ? DILX - Normal Termination HSJ> 7.3.2.
Example 7–3 (Cont.
Example 7–4 (Cont.
Example 7–5 (Cont.) Using Auto-Configure on an HSZ Controller HSZ> RUN DILX Disk Inline Exerciser - version 2.0 It is recommended that DILX only be run when there is no host activity present on the HSZ. Do you want to continue (y/n) [n]?y The Auto-Configure option will automatically select, for testing, all of the disk units configured. It will perform a very thorough test with *WRITES* enabled. The user will only be able to select the run time and performance summary options.
Example 7–5 (Cont.) Using Auto-Configure on an HSZ Controller HSZ> 7.3.2.6 Using the All Units Option on an HSZ40 Array Controller Example 7–6 shows a DILX run on an HSZ40 array controller without using the Auto-Configure option and choosing all units. Example 7–6 Using the All Units Option on an HSZ40 Controller HSZ> RUN DILX Disk Inline Exerciser - version 2.0 It is recommended that DILX only be run when there is no host activity present on the HSZ.
Example 7–6 (Cont.) Using the All Units Option on an HSZ40 Controller Write enable disk unit(s) to be tested (y/n) [n] ?y The write percentage will be set automatically.
Example 7–7 Using All Defaults on an HSZ40 Controller HSZ> RUN DILX Disk Inline Exerciser - version 2.0 It is recommended that DILX only be run when there is no host activity present on the HSZ. Do you want to continue (y/n) [n]?y The Auto-Configure option will automatically select, for testing, all of the disk units configured. It will perform a very thorough test with *WRITES* enabled. The user will only be able to select the run time and performance summary options.
7.3.3 TILX The tape inline exerciser (TILX) is a diagnostic tool used to exercise the data transfer capabilities of selected tape drives connected to an HSJ or HSD30 array controller. TILX exercises tapes in a way that simulates a high level of user activity. Thus, TILX can be used to determine the health of the controller and the tape drives connected to it. You can run TILX from the CLI, from VCS, or from a virtual (host) terminal.
7.3.3.2 Interrupting TILX Execution Use the following guidelines to interrupt TILX execution. Note The circumflex symbol (^) is equivalent to the Ctrl key. You must press and hold the Ctrl key and type the character key given. Note Do not use Ctrl/G from a VCS, because it causes VCS to terminate. VCS acts on the sequence, and the sequence never gets sent to TILX. Use Ctrl/T when interrupting TILX from a VCS. • Ctrl/G causes TILX to produce a performance summary.
7.3.4.1 Using All Functions In Example 7–8, all functions are chosen for TILX using a longer run time and higher record count than the default. The performance statistics and a performance summary are displayed every 10 minutes. In this case, TILX is invoked from a maintenance terminal. This example was run on an HSJ40 array controller. Example 7–8 Using All Functions HSJ> RUN TILX Tape Inline Exerciser - version 1.
Example 7–8 (Cont.) Using All Functions Unit 50 Total IO Requests 724 Read Count 3 Write Count 681 Reposition Count 3 Total KB xfer 6718 Read 10 Write 6707 No errors detected Unit 52 Total IO Requests 731 Read Count 3 Write Count 687 Reposition Count 3 Total KB xfer 6743 Read 10 Write 6733 No errors detected Reuse Parameters (stop, continue, restart, change_unit) [stop] ? TILX - Normal Termination HSJ> 7.3.4.2 Using All Defaults In Example 7–9, all defaults are chosen for TILX.
Example 7–9 (Cont.
7.3.5.1 How to Run VTDPY Only one VTDPY session can be run on each controller at one time. Prior to running VTDPY, be sure the terminal is set in NOWRAP mode. Otherwise, the top line of the display scrolls off of the screen. To initiate VTDPY from the maintenance terminal at the CLI> prompt, enter the following command: CLI> RUN VTDPY To initiate VTDPY from a virtual terminal, refer to Section 8.3.1. 7.3.5.
Table 7–2 (Cont.) VTDPY Commands Command String Function EXIT Terminates program (same as QUIT). INTERVAL Changes update interval. HELP Displays help message text. REFRESH Refreshes the current display. QUIT Terminates program (same as EXIT). UPDATE Updates screen display. The keywords in the command strings can be abbreviated to the minimum number of characters that are necessary to uniquely identify the keyword.
HSJ40 S/N: ZG33700938 SW: V25J HW: 00−00 03−FEB−1995 16:52:34 90.8% Idle 2656 KB/S 0 Rq/S Pr Name Stk/Max Typ Sta CPU% Node HSJA4 Port 25 Unit ASWC KB/S Rd% Wr% Cm% HT% 0 NULL 0/ 0 Rn 90.8 SysId 42001019F529 D0410 a^ r 0 0 0 0 0 2 RECON 10/ 1 FNC Bl 0.0 D0411 a^ r 0 0 0 0 0 3 HPT 40/ 4 FNC Bl 1.6 Path A Pkts Pkts/S D0413 a^ r 0 0 0 0 0 9 VTDPY 10/ 3 DUP Rn 0.2 RCV 203 20 D0414 a^ r 0 0 0 0 0 17 FMTHRD 10/ 1 FNC Bl 0.0 ACK 183 18 D0415 a^ r 0 0 0 0 0 18 DS_HB 10/ 1 FNC Bl 0.
HSD30 S/N: CX40300006 SW: V25D HW: 00−00 03−FEB−1995 16:48:41 19.7% Idle 0 KB/S 0 Rq/S Up: 0 0:50.27 Pr Name Stk/Max Typ Sta CPU% Node HSDD6 Port 6 Unit ASWC KB/S Rd% Wr% Cm% HT% Unit ASWC KB/S Rd% Wr% Cm% HT% 0 NULL 0/ 0 Rn 19.7 SysId 42001106E115 D2691 o^ r 0 0 0 0 0 2 RECON 10/ 1 FNC Bl 0.0 D2692 o^ r 0 0 0 0 0 3 HPT 40/ 8 FNC Rn 1.2 DSSI Pkts Pkts/S D2693 o^ r 0 0 0 0 0 8 VTDPY 10/ 3 DUP Rn 0.2 RCV 342 31 17 FMTHRD 10/ 1 FNC Bl 0.0 ACK 343 31 18 DS_HB 10/ 1 FNC Bl 0.
61.4% Idle Pr Name Stk/Max Typ Sta CPU% SCSI Target 2 0 NULL 0/ 0 Rn 61.4 2 RECON 10/ 1 FNC Bl 0.0 3 SHIS 40/ 7 FNC Rn 32.3 Xfer Rate 8 VTDPY 10/ 3 DUP Rn 0.2 Id Mhz 18 SCSIVT 10/ 1 FNC Bl 0.0 0 = Asynch 19 DS_HB 10/ 1 FNC Bl 0.0 1 = Asynch 24 VA 10/ 1 FNC Bl 0.0 2 = This 25 DS_1 40/ 6 FNC Bl 5.5 3 = Asynch 26 DS_0 20/ 1 FNC Bl 0.4 4 = Asynch 27 CLIMAIN 16/ 7 FNC Bl 0.0 5 = Asynch 28 NVFOC 10/ 1 FNC Bl 0.0 6 = 3.57 29 REMOTE 10/ 1 FNC Bl 0.0 7 = Asynch 30 FOC 20/ 2 FNC Bl 0.0 31 DUART 10/ 1 FNC Bl 0.
HSJ40 S/N: ZG33700938 SW: V25J HW: 00−00 03−FEB−1995 0.
HSJ40 S/N: ZG33700938 SW: V25J HW: 00−00 03−FEB−1995 16:53:26 80.1% Idle 2719 KB/S 23 Rq/S Up: 0 0:25.
Pr Name 0 NULL 2 RECON 3 HPT 9 VTDPY 17 FMTHRD 18 DS_HB 19 DUP 20 SCS 21 MSCP 23 VA 24 DS_1 25 DS_0 26 HIS 27 CLIMAIN 28 NVFOC 29 REMOTE 30 FOC 31 DUART 0.0% Idle 2717 KB/S Stk/Max Typ Sta CPU% Target 0/ 0 Rn 0.0 01234567 10/ 1 FNC Bl 0.0 P1DDFDDDhH 40/ 4 FNC Rn 100.0 o2DDDDDDhH 10/ 3 DUP Rn 0.0 r3DDDDDDhH 10/ 1 FNC Bl 0.0 t4DDDDDDhH 10/ 1 FNC Bl 0.0 5DDDDDDhH 10/ 1 FNC Bl 0.0 6DDDDDDhH 10/ 1 FNC Bl 0.0 20/ 1 FNC Bl 0.0 Connections 10/ 1 FNC Bl 0.0 0123456789 40/ 3 FNC Rn 0.0 0MVMMM..... 20/ 2 FNC Bl 0.
Pr Name 0 NULL 2 RECON 3 HPT 8 VTDPY 17 FMTHRD 18 DS_HB 19 DUP 20 SCS 21 MSCP 23 VA 24 DS_1 25 DS_0 26 HIS 27 CLIMAIN 28 NVFOC 29 REMOTE 30 FOC 31 DUART 0.0% Idle 0 KB/S Stk/Max Typ Sta CPU% Target 0/ 0 Rn 0.0 01234567 10/ 1 FNC Bl 0.0 P1 DDDDD H 40/ 8 FNC Rn 0.0 o2 DDDDD H 10/ 3 DUP Rn 100.0 r3 DDDDD H 10/ 1 FNC Bl 0.0 t 10/ 1 FNC Bl 0.0 10/ 1 FNC Bl 0.0 10/ 1 FNC Bl 0.0 20/ 1 FNC Bl 0.0 Connections 10/ 1 FNC Bl 0.0 0123456789 40/ 11 FNC Rn 0.0 0.......M 20/ 1 FNC Rn 0.0 1 10/ 1 FNC Bl 0.
60.9% Idle Pr Name Stk/Max Typ Sta 0 NULL 0/ 0 Rn 2 RECON 10/ 1 FNC Bl 3 SHIS 40/ 7 FNC Rn 8 VTDPY 10/ 3 DUP Rn 18 SCSIVT 10/ 1 FNC Bl 19 DS_HB 10/ 1 FNC Bl 24 VA 10/ 1 FNC Bl 25 DS_1 40/ 6 FNC Bl 26 DS_0 20/ 1 FNC Bl 27 CLIMAIN 16/ 7 FNC Bl 28 NVFOC 10/ 1 FNC Bl 29 REMOTE 10/ 1 FNC Bl 30 FOC 20/ 2 FNC Bl 31 DUART 10/ 1 FNC Bl 861 KB/S 293 Rq/S CPU% Target Unit ASWC 60.9 01234567 D0000 o^ b 0.0 P1D D D H D0001 o^ b 32.7 o2 D D DDH D0002 o^ b 0.1 r3D D D H D0003 o^ b 0.0 t4 D D DH D0004 o^ b 0.
Display Header ! % HSJ40 S/N: CX00000002 VTDPY Monitor " SW: V25J # HW: A-02 $ Description This subdisplay provides title information for the display. For 132 column displays, this subdisplay is displayed across one line.
Date and Time ! " 29-JAN-1995 13:46:34 Up: 1 3:45.19 Description This subdisplay provides time information for the display. ! " System date and time. This information is not displayed for SCSI based HS array controllers. Time in days, hours, minutes, and seconds since the last controller boot.
Controller Performance Summary 47.2% Idle ! 1225 KB/S "106 Rq/S # Description This subdisplay provides total system performance information. ! " # Policy processor idle rate. Cumulative data transfer rate in kilobytes per second. When logical units are being displayed, this is the transfer rate between the host and the controller. When physical devices are being displayed, this is the transfer rate between the controller and the devices. Cumulative unit or device request rate per second.
Controller Threads Display ! " # Typ$ Sta% CPU%& Pr Name Stk/Max 0 NULL 0/ 0 3 HPT 40/ 7 8 VTDPY 10/ 3 18 FMTHRD 10/ 2 19 DS_HB 10/ 2 20 DUP 10/ 2 21 SCS 10/ 2 22 MSCP 20/ 6 24 VA 10/ 3 25 DS_1 40/ 6 26 DS_0 20/ 4 27 HIS 10/ 2 28 CLIMAIN 16/ 6 30 FOC 16/ 4 31 DUART 10/ 2 FNC DUP FNC FNC FNC FNC FNC FNC FNC FNC FNC FNC FNC FNC Rn Rn Rn Bl Bl Bl Bl Bl Bl Rn Bl Bl Bl Bl Bl 47.2 40.3 0.1 0.0 0.0 1.3 0.0 0.0 1.2 8.9 0.0 0.0 0.0 0.0 0.
& The CPU% column lists the percentage of execution time credited to each thread since the last screen update. The values may not add up to exactly 100 percent due to both rounding errors and the fact that there may not be enough room to display all of the threads. An unexpected amount of time may be credited to some threads because the controller’s firmware architecture allows code from one thread to execute in the context of another thread without a context switch.
Table 7–3 (Cont.) Thread Description Thread Name Description RMGR The thread that manages the data buffer pool. RECON The thread that rebuilds the parity blocks on RAID level 5 storagesets when needed and manages mirrorset copy operations when necessary. SCS The SCS directory thread. SCSIVT A thread that provides a virtual terminal connection to the CLI over the host SCSI bus. SHIS The host SCSI protocol interface thread for SCSI controllers. TILX A local program that exercises tape devices.
CI/DSSI Host Port Characteristics ! # Node HSJ501 Port 13 SysId 4200100D0720 " Description This subdisplay shows the current host port identification information. This subdisplay is available only for CI or DSSI based controllers.
SCSI Host Port Characteristics !"# $ % Xfer Rate T W I Mhz 1 W 7 10.00 2 W Async Description This subdisplay shows the current host port SCSI target identification, any initiator which has negotiated synchronous transfers, and the negotiated transfer method currently in use between the controller and the initiators. This subdisplay is available only for SCSI based HS array controllers. ! " # $ % SCSI host port target ID. Transfer width. W indicates 16 bit or wide transfers are being used.
CI Performance Display Path A Pkts Pkts/S RCV 5710 519 ACK 11805 1073 NAK 2073 188 NOR 1072 97 ! " # $ Path B Pkts Pkts/S RCV 5869 533 ACK 11318 1028 NAK 2164 196 NOR 445 40 Description This display indicates the number of packets sent and received over each CI path and the packet rate. This display is available only on CI based controllers.
DSSI Performance Display DSSI RCV ACK NAK NOR Pkts Pkts/S 5710 519 11805 1073 2073 188 1072 97 ! " # $ Description This display indicates the number of packets sent and received through the DSSI port and the packet rate. This display is available only on DSSI based controllers.
CI/DSSI Connection Status Connections 0123456789 0........MM 1..C.MV.... 2.......... 3.. ! " Description This display shows the current status of any connections to a remote CI or DSSI node. This display is available only on CI and DSSI based controllers. ! " Each position in the data field represents one of the possible nodes to which the controller can communicate.
CI/DSSI Host Path Status Path Status 0123456789 0........^^ 1..A.B^.... 2.......X.. 3.. ! " Description This display indicates the path status to any system for which a virtual circuit exists. This display is available only on CI and DSSI based controllers. ! " Each position in the data field represents one of the possible nodes to which the controller can communicate.
Device SCSI Status Target 01234567 P1 DDDDFhH o2TTT T hH r3DDD hH t4DDDDDDhH 5DDDD hH 6 hH ! " # Description This display shows what devices the controller has been able to identify on the device busses. Note The controller does not look for devices that are not configured into the nonvolatile memory using the CLI ADD command. ! " # The column headings indicate the SCSI target numbers for the devices. SCSI targets are in the range 0 through 7. Target 7 is always used by a controller.
Unit Status (abbreviated) ! ASWC" KB/S# Rd%$ Wr%% Cm%& HT%' Unit D0110 D0120 D0130 T0220 T0230 a^ r a^ r o^ r av o^ 0 0 236 0 123 0 0 100 0 0 0 0 0 0 100 0 0 0 0 0 0 0 100 0 0 Description This subdisplay shows the status of the logical units that are known to the controller firmware. It also indicates performance information for the units. Up to 42 units can be displayed in this subdisplay.
• A space in this column indicates the availability is unknown. The spindle state is indicated using the following characters: • ^—For disks, this symbol indicates the device is at speed. For tapes, it indicates the tape is loaded. • >—For disks, this symbol indicates the device is spinning up. For tapes, it indicates the tape is loading. • <—For disks, this symbol indicates the device is spinning down. For tapes, it indicates the tape is unloading.
Unit Status (full) ! ASWC" Unit D0003 D0250 D0251 D0262 D0280 D0351 D0911 D1000 o^ o^ o^ a^ o^ a^ a^ a^ r r r r r r r r # Rd%$ KB/S 382 382 284 0 497 0 0 0 0 100 100 0 44 0 0 0 % Cm%& HT%' PH%( MS%) Purge+> BlChd+? BlHit+@ Wr% 100 0 0 0 55 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100 100 0 100 0 0 0 0 0 0 0 0 0 0 0 6880 6880 5120 0 9011 0 0 0 0 0 0 0 0 0 0 0 Description This subdisplay shows the status of the logical units that are known to the controller firmware.
• v—Offline, No Volume Mounted. The device does not contain media. • x—Online to other controller. Not available for use by this controller. • A space in this column indicates the availability is unknown. The spindle state is indicated using the following characters: • ^—For disks, this symbol indicates the device is at speed. For tapes, it indicates the tape is loaded. • >—For disks, this symbol indicates the device is spinning up. For tapes, it indicates the tape is loading.
+? +@ BlChd—This column shows the number of blocks added to the cache in the last update interval. BlHit—This column shows the number of cached data blocks ‘‘hit’’ in the last update interval.
Device Status ! ASWF" Rq/S# RdKB/S$ WrKB/S% Que& Tg' CR( BR) TR+> PTL D100 D120 D140 D210 D230 D300 D310 D320 D400 D410 D420 D430 D440 D450 D500 D510 D520 D530 A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ A^ 0 0 0 11 0 11 0 36 11 0 36 0 0 0 11 0 0 47 0 0 0 93 0 93 0 247 93 0 247 0 0 0 93 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 375 11 0 0 1 0 2 0 12 2 0 10 0 0 0 1 0 0 6 0 0 0 1 0 1 0 10 1 0 8 0 0 0 1 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The spindle state is indicated using the following characters: • ^—For disks, this symbol indicates the device is at speed. For tapes, it indicates the tape is loaded. • >—For disks, this symbol indicates the device is spinning up. For tapes, it indicates the tape is loading. • <—For disks, this symbol indicates the device is spinning down. For tapes, it indicates the tape is unloading. • v—For disks, this symbol indicates the device is stopped. For tapes, it indicates the tape is unloaded.
Device SCSI Port Performance ! Rq/S" RdKB/S# WrKB/S$ CR% BR& TR' Port 1 2 3 4 5 6 0 11 48 48 58 0 0 93 341 340 93 0 0 0 0 0 375 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Description This subdisplay shows the accumulated I/O performance values and bus statistics for the SCSI device ports. The subdisplay for a controller that has six SCSI device ports in shown. ! " # $ % & ' The Port column indicates the number of the SCSI device port.
Help Example VTDPY> HELP Available VTDPY commands: ^C - Prompt for commands ^G or ^Z - Update screen ^O - Pause/Resume screen updates ^Y - Terminate program ^R or ^W - Refresh screen DISPLAY CACHE - Use 132 column unit caching statistics display DISPLAY DEFAULT - Use default 132 column system performance display DISPLAY DEVICE - Use 132 column device performance display DISPLAY STATUS - Use 80 column controller status display EXIT - Terminate program (same as QUIT) INTERVAL - Change update interva
7.3.6 Firmware Licensing System (FLS) The firmware licensing system (FLS) enables or disables the licensed value-added software features (RAID 3/5, write-back cache, and mirroring) of the HS array controller. You may use the FLS utility to perform the following tasks: • Enable or disable optional functions for your controller • Try an optional feature before purchasing the license to use it • Change your license key for an option Start FLS from the CLI prompt.
The following example shows the FLS main menu: CLI> RUN FLS -------------------------------------------------------------------------------Firmware Licensing System (FLS) on node BERT State License Key Option -------------------------------------------------------------------------------RAID DISABLED INVALID *none* WBCA ENABLED ******INVALID!****** *none* MIRR ENABLED VALID ACME_WIDGET_CORP........
1. Enter new license key+CRC for RAID 2. Enter new license key+CRC for WBCA 3. Enter new license key+CRC for MIRR 0.
$ This entry in the FLS display shows that write-back cache is enabled under a valid license. 7.3.6.6 Messages This section lists the messages that you may receive from FLS. option has been turned on without a valid license Explanation: You have activated the option named by option without entering a valid license key. You can evaluate this option for a time to determine its value, and you will receive a valid license key when you purchase the license for the option.
Before running the Configure Utility, use the CLI> SHOW DEVICES command to verify the list of devices that are currently connected to the HS array controller, as shown in the following example. The example shows the Configure Utility as it is run on an HSJ or HSD30 array controller. The text of the prompts may change slightly when run on other controllers from the HS array controller family. Note Program names are limited to six characters. Therefore, to run this program, use RUN CONFIG as your command.
7.3.8 HSZterm Utility The HSZterm utility provides a virtual terminal facility and a command passing capability for communicating with an HSZ controller over the SCSI bus. This utility allows DEC OSF/1 system administrators to configure HSZ array controllers using terminal windows. The window behaves as if it were a terminal attached directly to the HSZ controller. All controller commands except those that make use of cursor positioning escape sequences are supported.
This section presents an overview of one configuration, from start to finish. Only one example is described because of the many combinations of choices that can be made during any configuring situation. Note that the options for passthrough containers are not available for HSZ array controllers. 7.3.10.1 Restrictions The following restrictions apply to CFMENU: • For dual-redundant configurations, do not run CFMENU on both controllers at the same time.
Figure 7–10 CFMENU Device Menu (Before Adding Devices) ------------------------------------------------ CFMENU Configuration Menu Utility -----------------------------------------------DEVICE MENU: |Unconfig'd| Config'd Device Product Stor.set Stor.set Chnk Trn In- ReW W 1. Add a device from list | Dev.PTLs | PTLs Name ID Name Typ/Sz Size sp. it'd duc Unit P B of PTLs not configured |----------| ------ --------- ---------------- --------- ------- ------ --- ---- --- ----- - (marked with ^) |^110 (dsk)| 2.
• Init’d—Initialized. This field identifies whether or not the device is initialized. After adding devices, return to the main menu. 7.3.10.4 Adding Mirrorsets Enter option 2 from the main menu to work with mirrorsets. From the mirrorset menu (see Figure 7–12), enter option 1 to add a mirrorset. CFMENU prompts you for how many and which devices from the configured PTLs list you wish to include in the mirrorset (2 to 14 devices allowable).
Figure 7–13 CFMENU Stripeset Menu ------------------------------------------------ CFMENU Configuration Menu Utility -----------------------------------------------STRIPESET MENU: |Unconfig'd| Config'd Device Product Stor.set Stor.set Chnk Trn In- ReW W 1. Create a stripeset | Dev.PTLs | PTLs Name ID Name Typ/Sz Size sp. it'd duc Unit P B (eligible entities marked |----------| ------ --------- ---------------- --------- ------- ------ --- ---- --- ----- - by ^) | |disks: 220 ^ DISK220 RZ25 (C) DEC N Y 2.
CFMENU updates the following fields after adding a RAIDset: • Stor.set Name—Storageset name. CFMENU automatically assigns the name ‘‘Rx’’ to a RAIDset. You cannot alter this automatic CFMENU naming convention. • Stor.set Typ/Sz—Storageset type. This field will read ‘‘RAD’’ for storagesets that are RAIDsets, and indicate the number of members in the RAIDset. • Chnk Size—Chunk size. This field will read ‘‘unk’’ (unknown) until you initialize the RAIDset. 7.3.10.
7.3.10.8 Adding Passthroughs (HSJ and HSD30 array controllers) Enter option 5 from the main menu to work with passthroughs. From the passthrough menu (see Figure 7–16), enter option 1 to add a passthrough. CFMENU will ask you which device from the configured PTLs list you wish to include in the passthrough. Although you may add any generic SCSI device, you will normally add devices requiring the use of MSCP read and write operations as a transport for SCSI commands (such as ‘‘jukebox’’ style loaders).
7.3.10.9 Initializing Containers Enter option 6 (HSJ and HSD30 array controllers) or option 5 (HSZ array controllers) from the main menu to initialize containers (devices or storagesets). From the initialization menu (see Figure 7–17), enter option 1. CFMENU will ask you if you want to initialize each eligible container. In addition, CFMENU will ask you to decide on other operating qualifiers, depending on whether the container is a device, mirrorset, stripeset, or RAIDset.
7.3.10.10 Adding Units Enter option 7 (HSJ and HSD30 array controllers) or option 6 (HSZ array controllers) from the main menu to work with units. From the unit menu (see Figure 7–18), enter option 1 to add a unit. CFMENU will ask you which initialized containers you wish to create units from. CFMENU also will ask you to assign a unit number. (The program automatically assigns a ‘‘D’’ or ‘‘T’’ to the unit number when listing the unit, as shown in Figure 7–18.
7.3.10.11 Terminal Setup You can enter option 8 (HSJ and HSD30 array controllers) or option 7 (HSZ array controllers) from the main menu to set the number of rows CFMENU will display. This feature is available primarily for terminals with the capability of displaying more than 24 rows. 7.3.10.12 Messages This section lists the messages, other than the standard CLI messages, that CFMENU will display. However, most messages you will see will be those sent by the CLI, and are described in Appendix B.
There are no devices available to use as a replacement. Explanation: You chose to replace a member of a reduced RAIDset, but there are no disks eligible to use for the replacement member. Eligible disks must have the NOTRANSPORTABLE switch set and may not be part of any higher-level configuration such as units, storagesets, or the spareset or failedset.
There are no entities eligible for initialization. Explanation: You chose to initialize a device or storageset, but there are currently no devices or storagesets that are eligible to be initialized. Tapes, CD–ROMs, loaders, and passthrough devices may not be initialized. Optical memory devices, stripesets, and RAIDsets may be initialized, but only if they are not already configured as a unit.
Waiting for completion of CLI command... Explanation: Some CLI commands take a long time to complete, such as initializing a large RAIDset or adding a tape drive that needs to have the tape rewound. CFMENU prints out this message periodically to inform you that it is still waiting for the last CLI command to finish. 7.3.10.13 Exiting CFMENU Enter the last option from the main menu to stop CFMENU and return to the CLI. (You also may enter Ctrl/C or Ctrl/Y to abort CFMENU.) 7.3.
7.3.11.2 Code Loading The Code Load option of the CLCP utility is invoked from an external processor (typically a personal computer) connected to the maintenance terminal port on the front bezel of the controller. The processor must be configured to run the KERMIT terminal protocol at 19,200 baud, with 8 data bits, no parity, and one stop bit. When running the Code Load option of the CLCP utility from a personal computer, all screen savers should be disabled.
Perform the following steps before continuing: * get new image file on serial line host computer * configure KERMIT with the following parameters: terminal speed 19200 baud, eight bit, no parity, 1 stop bit WARNING: proceeding with Code Load will overwrite the current content of your program card with a new image. Enter Y (then RETURN) to continue [N]: ? Y Start KERMIT now... 3.
Code Load Aborted by Interrupting KERMIT Before the Code Image Is Sent CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..2) [0] ? 1 ---------------------------------------------------------------------You have selected the Code Load local program. This program is used to load a new firmware image on the program card currently inserted in the controller.
2. Invoke the program as follows: CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..2) [0] ? 2 ---------------------------------------------------------------------Code Patch Main Menu 0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..3) [0] ? 3.
List of Current Patches The List Patches option allows you to display a listing of controller firmware versions, and the currently installed patches that apply to them. The following is an example of the List Patches option: CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..
Deleting a Patch and a Firmware Version The Delete Patches option allows you to remove previously installed patches from controller non-volatile memory. The program displays the currently installed patches and patches to be deleted. The Code Patch option of the CLCP utility verifies that the patch requested for deletion exists, and that it is not a dependent patch for a higher numbered installed patch. It allows you to delete only one patch at a time.
Code Patch Main Menu 0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..3) [0] ? 3 The following patches are currently stored in the patch area: Firmware Version - Patch number(s) ______ _________ XLZ6 2,1 Currently, 94% of the patch area is free. Code Patch Main Menu 0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..
Currently, 94% of the patch area is free. Firmware Version of patch to delete ? asdf Firmware Version ASDF does not have any patches to delete Code Patch Main Menu 0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..
0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..3) [0] ? Attempting to Enter a Patch Number That Already Exists The following is an example of what happens when you enter a patch number that is already loaded for the specified firmware version: CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..
Enter option number (0..3) [0] ? Attempting to Enter a Patch When Dependent Patches Are Not Present The following is an example of what happens when you try to enter a patch for a firmware version, and dependent (lower numbered) patches have not been previously installed: CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..
CLI> RUN CLCP Select an option from the following list: Code Load & Code Patch local program Main Menu 0: Exit 1: Enter Code LOAD local program 2: Enter Code PATCH local program Enter option number (0..2) [0] ? 2 ---------------------------------------------------------------------You have selected the Code Patch local program. This program is used to manage firmware code patches. Select an option from the following list: Type ^Y or ^C (then RETURN) at any time to abort Code Patch.
Type ^Y or ^C (then RETURN) at any time to abort Code Patch. Type ^Z (then RETURN) at any prompt to choose the default for the remaining questions. Version: [V15J ]? WARNING The patch you are entering is not for the current firmware version (XLZ6 ). The patch you are entering is for firmware version V15J . This patch will NOT be applied to the current firmware.
This is the Delete Patches option. The program prompts you for the firmware version and patch number for deletion. If you select a patch for deletion that is required for another patch, all dependent patches are also selected for deletion. The program lists your deletion selections and asks if you wish to continue. Type ^Y or ^C (then RETURN) at any time to abort Code Patch.
0: 1: 2: 3: Exit Enter a Patch Delete Patches List Patches Enter option number (0..3) [0] ? 1 This is the Enter a Patch option. The program prompts you for the patch information, one line at time. Be careful to enter the information exactly as it appears on the patch release. Patches may be installed for any version of firmware; however, patches entered for firmware versions other than XLZ6 are not applied until the matching version of firmware is installed.
Although the operations involved in most data snapshot processes are commonplace, the reaction in response to failures during the operation is heavily influenced by each installation’s operational needs. For this reason, CLONE does not attempt to automate the error recovery process. If you encounter an error during a CLONE operation, you will need to resolve the situation by manually issuing the appropriate CLI commands. The following circumstance will cause CLONE to cease operation: 1.
The new mirrorset level only affects clones of single-disk units and stripesets. Mirrorsets and striped mirrorsets already use mirroring, and any clones from such units will have the same structure as the original units. The following are examples for cloning a single device unit, cloning a stripeset, cloning a mirrorset, and finally, cloning a striped mirrorset.
5. SET mirrorset-container-name MEMBERSHIP= number-of-members 6. SET mirrorset-container-name REPLACE= disk-device-name Wait for member NORMALIZATION 7. REDUCE disk-device-name 8. UNMIRROR disk-device-name 9. ADD MIRRORSET mirrorset-container-name disk-device-name 10. INITIALIZE container-name NODESTROY 11.
Example 7–11 Cloning a Stripeset HSJ> SHOW STORAGESETS Name Storageset Uses Used by -----------------------------------------------------------------------------ST1 stripeset DISK130 DISK200 D799 HSJ> run clone Clone Local Program Invoked Units available for cloning: 110 799 Enter unit to clone ? 799 Clone will create a new unit which is a copy of unit 799. Enter the unit number which you want assigned to the new unit ? 798 The new unit may be added using one of the following methods: 1.
Example 7–11 (Cont.
The following is an example of the manual commands for cloning a stripeset: HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> ! " # $ SHOW DEVICES SHOW STORAGESETS MIRROR DISK130 C_MA SET C_MA NOPOLICY SET C_MA MEMBERSHIP=2 SET C_MA REPLACE=DISK220 MIRROR DISK200 C_MB SET C_MB NOPOLICY SET C_MB MEMBERSHIP=2 SET C_MB REPLACE=DISK240 ( % ' ) & +> ...wait for NORMALIZATION OF DISK220 and DISK240...
+D +E +F +G ,> Creates a stripeset called C_ST1 from the two newly-created mirrorsets C_MA and C_MB. Initializes stripeset C_ST1 with a chunksize of 128 blocks, but does not write new metadata onto the members. Creates a logical unit for the host from the initialized stripeset C_ST1 and names the unit D798. Shows the device configuration after the clone is complete. Shows the stripesets after the clone is complete.
Example 7–12 (Cont.) Cloning a Mirrorset Clone Local Program Invoked Units available for cloning: 411 Enter unit to clone ?411 Enter the unit number which you want assigned to the new unit?499 The new unit may be added using one of the following methods: 1. Clone will stall after all members have been copied. The user must then hit RETURN to cause the new unit to be added. 2. After all members have been copied, the unit will be added automatically.
Example 7–12 (Cont.) Cloning a Mirrorset Name Storageset Uses Used by -----------------------------------------------------------------------------C_M1 mirrorset DISK310 D499 M1 mirrorset DISK110 DISK220 D411 HSJ> To manually clone a mirrorset use the following commands in sequential order: 1. SHOW DEVICES 2. SHOW MIRRORSETS 3. SET mirrorset-container-name NOPOLICY 4. SET mirrorset-container-name MEMBERSHIP=3 5. SET mirrorset-container-name REPLACE= disk-device-name Wait for NORMALIZATION. 6.
& ' ( ) +> +? Removes DISK310 from mirrorset M1 after the copy operation has completed. Creates a mirrorset called C_M1 from DISK310. Initializes mirrorset C_M1 with a chunksize of 128 blocks, but does not write new metadata onto the members. Creates a logical unit for the host from mirrorset C_M1 and names the unit D499. Shows the device configuration after the clone is complete. Shows the mirrorsets after the clone is complete.
Example 7–13 Cloning a Striped Mirrorset HSJ> SHOW DEVICES Name Type Port Targ Lun Used by -----------------------------------------------------------------------------DISK110 DISK150 DISK220 DISK310 DISK350 DISK420 DISK510 DISK550 DISK620 disk disk disk disk disk disk disk disk disk 1 1 2 3 3 4 5 5 6 1 5 2 1 5 2 1 5 2 0 0 0 0 0 0 0 0 0 M1 M1 M2 M2 M3 M3 HSJ> SHOW MIRRORSETS Name Storageset Uses Used by -----------------------------------------------------------------------------M1 mirrorset DISK110
Example 7–13 (Cont.) Cloning a Striped Mirrorset Clone Local Program Invoked Units available for cloning: 411 Enter unit to clone ?411 Clone will create a new unit which is a copy of unit 411. Enter the unit number which you want assigned to the new unit?499 The new unit may be added using one of the following methods: 1. Clone will stall after all members have been copied. The user must then hit RETURN to cause the new unit to be added. 2.
Example 7–13 (Cont.
To manually clone a striped mirrorset use the following commands in sequential order: 1. SHOW DEVICES 2. SHOW MIRRORSETS 3. SHOW STORAGESETS 4. SET mirrorset-container-name1 NOPOLICY 5. SET mirrorset-container-name1 MEMBERSHIP= 6. SET mirrorset-container-name1 REPLACE= 7. SET mirrorset-container-name2 NOPOLICY 8. SET mirrorset-container-name2 MEMBERSHIP= 9. SET mirrorset-container-name2 REPLACE= 10. SET mirrorset-container-name3 NOPOLICY 11. SET mirrorset-container-name3 MEMBERSHIP= 12.
HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> HSJ> ! " # $ % & ' ( ) +> +? +@ +A +B +C +D +E +F +G +B +C +D +F +A REDUCE DISK350 DISK150 DISK550 ADD MIRRORSET C_M1 DISK350 ADD MIRRORSET C_M2 DISK150 ADD MIRRORSET C_M3 DISK550 ADD STRIPESET C_ST1 C_M1 C_M2 C_M3 INITIALIZE C_ST1 NODESTROY ADD UNIT D499 C_ST1 SHOW DEVICES SHOW MIRRORSETS SHOW STORAGESETS ,> ,? ,@ +G +E Displays the devices available for cloning. Displays the configured mirrorsets. Displays the configured storagesets.
,> ,? ,@ Shows the device configuration after the clone is complete. Shows the mirrorsets after the clone is complete. Shows the storagesets after the clone is complete. 7.3.13 Command Disks (HSJ and HSD30 Array Controllers) A special storage configuration is available for communication with generic SCSI devices. The concept centers around a virtual unit, called a command disk, that allows an MSCP-based host to control any physical SCSI device.
4. Add a passthrough container, the bridge between the physical loader and the virtual unit. 5. Add the passthrough container as a virtual unit visible to the host. The following example shows the CLI commands necessary to add a tape library device to your configuration using a command disk. Note that the PTL addresses for devices in your storage subsystem will probably be different.
Beginning installation of JUDRIVER V1.4 at 15:54 . . . For each command disk on your system, enter the DU device designation. Type CTRL/Z when there are no more command disks. . . . * Enter the first DU device designation: $6$dua622 * Enter the next DU device or CTRL/Z if done: Exit %JUDRIVER-I-DEFINE, defining device JUA0 %JUDRIVER-I-LINK, Linking JUDRIVER ... %JUDRIVER-I-LINK, Linking JUDRIVER_INIT ... " ! " ! In this case, the loader mechanism resides under unit number D622.
8 Operating System Support 8.
8.2 HS Array Controller System Management System management of the HS array controllers is meant to be done remotely over a VCS or DUP connection (or by using HSZterm for HSZ array controllers) with the exception of entering the vital controller parameters (controller ID and so forth). Only the vital controller parameters must be entered from a maintenance terminal connected directly to the maintenance terminal port at initial installation or when replacing a failed controller.
Creating a LOG File If you do not want to have a log file created, delete the /LOG= portion of the command. Substitute the name of the program you want to run after the /TASK= portion of the command, such as DILX if you want to run a test without going into the controller’s CLI. Using the /LOG= qualifier creates a log file of your CLI session when you create your HS array controller’s configuration. You are instructed to log out of the session and print the file after the configuration is defined.
Device Behavior After Controller Reinitialization When a controller reinitialization occurs in a multiple host VAXcluster system, the hosts automatically reconnect to the devices attached to the controller. With OpenVMS operating system versions (both VAX and Alpha), these reconnections may be via another host. This behavior is not unique to the HSJ array controller and may be changed in a future version of the OpenVMS operating system. Maximum Byte Count for ERASE Commands As of HSOF Version 1.
DEC 7000 and DEC 10000 Systems Disks attached to HSJ array controllers may be used as initialization devices for DEC 7000 or DEC 10000 processors provided the following condition is met: • DEC 7000 or DEC 10000 console code must be at Version 3.1 or higher. Note Currently, HSD30 array controllers are not supported under DEC 7000/10000 systems. 8.3.2.2 HSJ and HSD30 Array Controller-Attached Disk Drives and VMS AUTOGEN Program The OpenVMS AUTOGEN.
Completing this procedure causes HSJ or HSD30 controller-attached disk drives to be recognized as supported device types. For the OpenVMS VAX Version 6.0 operating system, the AUTOGEN.COM DCL procedure does not support device types above 137 although HSX00 and HSX01 are properly defined in the speed list. To circumvent this problem, perform the following steps: 1. Make a copy of the AUTOGEN.COM DCL file in case restoration of the original state is required. 2. Edit the AUTOGEN.COM file.
$ SHOW DEVICE $3$DUA1/FULL Disk $3$DUA1: (FRED), device type (type not yet identified), is online, fileoriented device, shareable, available to cluster, error logging is enabled.
The maximum size you can specify for a volume is one-hundredth the size of the volume; the minimum size you can specify is calculated with the following formula: disk_size (number_of_blocks) ___________________________ 255*4096 For Files–11 On-Disk Structure Level 2 disks, the cluster size default depends on the disk capacity; disks that are 50,000 blocks or larger have a default cluster size of 3, those smaller than 50,000 blocks have a default value of 1.
• Physical device. The physical device type must be identical for all devices to be used in a single host-based shadow set. The physical device type is obtained by entering the controller SHOW DISK command. Storagesets created in the controller’s configuration and combined with other storagesets in a host-based shadow set also must be identical. 8.3.
Example 8–1 OpenVMS Alpha V6.1 ERF Error Log Report V M S SYSTEM ERROR REPORT *************************** ENTRY ERROR SEQUENCE 5. DATE/TIME 28-APR-1994 11:39:40.33 SYSTEM UPTIME: 0 DAYS 00:01:41 SCS NODE: MTX2 COMPILED 9-AUG-1994 13:41:37 PAGE 758. 1122. ******************************* LOGGED ON: CPU_TYPE 00000002 SYS_TYPE 00000003 OpenVMS Alpha X6.1-FT7 HW_MODEL: 00000401 Hardware Model = 1025.
Example 8–1 (Cont.) OpenVMS Alpha V6.1 ERF Error Log Report CONTROLLER DEPENDENT INFORMATION LONGWORD 1. 030C4002 LONGWORD 2. 00003C51 LONGWORD 3. 00000000 LONGWORD 4. 000B9331 LONGWORD 5. 00000000 LONGWORD 6. 00000000 LONGWORD 7. 00000000 LONGWORD 8. 00000000 LONGWORD 9. 1F000504 LONGWORD 10. 36325A52 LONGWORD 11. 20202020 LONGWORD 12. 29432820 LONGWORD 13. 43454420 LONGWORD 14. 34333533 LONGWORD 15. 37313739 LONGWORD 16. 00000000 LONGWORD 17. 00000004 LONGWORD 18.
Some subcodes and all controller-dependent information is not decoded by ERF. Instructions for how to decode this information is contained in the StorageWorks Controllers HS Family of Array Controllers Service Manual. Example 8–2 is an example of an OpenVMS VAX error log that did not decode the subcodes and the controller dependent information.
Refer to the StorageWorks Array Controllers HS Family of Array Controllers Service Manual for ERF decode and error packet information for the HSJ or HSD30 array controllers. 8.3.8 DECevent for OpenVMS VAX and OpenVMS Alpha Operating Systems The DECevent utility is an event management utility that provides the interface between you and the operating system’s event logger.
On OpenVMS VAX and OpenVMS Alpha systems, you need the SYSPRV privilege to use the DECevent utility. You also need the DIAGNOSE privilege to enable continuous display (the /CONTINUOUS qualifier) on the terminal. DECevent Version 1.1 comes bundled with the OpenVMS Alpha (Version 6.1 and higher) operating system as a customer installable option. The ERF (error report formatter) also comes bundled with the operating system as the primary service tool.
Example 8–3 (Cont.) DECevent Bit-to-Text Translation Error Event Report MSCP Unique Controller-ID x0000000940802576 MSCP Controller Model 40. HSJ40 HS Array Controller MSCP Controller Class 1. Mass Storage Controller class Controller SW version 20. Controller HW version 73. MSCP MSCP MSCP Unit Unit Unique Unit-ID Unit Model Unit Class SW version HW version MSCP Event Code Multiunit code Error recovery Level Retry count Volume Serial Number Header code x0000000000000022 -1.
Example 8–3 (Cont.
8.4 DEC OSF/1 Support Refer to the DEC OSF/1 operating system documentation for operating system specific information. Refer to HS array controller-specific SPDs and release notes for supported versions. Not all controller models are currently supported by the DEC OSF/1 operating system. Note All configuration set up, parameter definitions, and utilities must be run by using an appropriate terminal emulator with a maintenance terminal connected to your controller’s terminal port.
Example 8–4 uerf HSZ40 Array Controller Error Event Log ----- EVENT INFORMATION ----EVENT CLASS OS EVENT TYPE SEQUENCE NUMBER OPERATING SYSTEM OCCURRED/LOGGED ON OCCURRED ON SYSTEM SYSTEM ID 199. 19.
Example 8–4 (Cont.) uerf HSZ40 Array Controller Error Event Log *MY ADDR CCB LENGTH FUNC CODE CAM_STATUS x8A960728 x00C0 x01 PATH ID TARGET ID TARGET LUN CAM FLAGS 14. 2. 2.
Example 8–4 (Cont.
Example 8–5 (Cont.) uerf Utility Error Event Using the -Z Qualifier ROUTINE NAME spo_process_ccb ----- CAM STRING ----A SCSI bus reset has been done ----- UNSUPPORTED ENTRY ----CAM ENTRY x00000436 RECORD ENTRY DUMP: RECORD HEADER 0000: 00040220 0010: 70737661 0020: 00000001 0030: FFFFFFFF 00000009 00060101 2F29198A 00003667 00000000 00000000 00000000 003300C7 FFFF0002 00000000 * .............)/* *joey..........* *..........3.....* *........ * RECORD BODY 0038: 000000C7 00000000 00000333 00000033 *...
information. Refer to the HSZ array controller release notes for supported host adapters and DEC OSF/1 version levels for ASE. 8.4.3 Configurations and Device Support for the HSZ Array Controllers The following sections describe the type of device support and how those devices can be used for the Version 2.5 firmware release for HSZ array controllers. Only disk-type devices (rzxx and ezxx) are supported at the time of printing.
The above naming format is recommended for the naming of DEC OSF/1 HSZ40 devices. This format prevents DEC OSF/1 naming conflicts between HSZ40 devices and other SCSI devices that may be present on the host system. The DEC OSF/1 utility that is used to create device special files is mknod. The user must be superuser to use this command. The default working directory must be /dev. Refer to the MAN pages on the mknod utility.
Examples 8–6 and 8–7 are examples for creating device special files. Example 8–6 Creating Device Special Files (NATIVE SCSI Host Adapter) Suppose you want to communicate with LUN 0 on an HSZ40 array controller at SCSI target ID 2 connected via the SCSI bus to a native SCSI host adapter at bus 1, and the HSZ40 unit designator is D200. The following steps create the device special files for this configuration, using the previously mentioned naming format of rzxnny: 1.
Example 8–7 Creating Device Special Files (SCSI Host Adapter) Suppose you want to communicate with LUN 4 on an HSZ40 array controller at SCSI target ID 2 connected via the SCSI bus to a KZTSA SCSI host adapter at bus 14, and the HSZ40 unit designator is D204. The following steps create the device special files for this configuration using the name format rzxnny: 1. Generate the file name for partition device special files: nn = (14 * 8) + 2 = 114 file name = rze114a (block) and rrze114a (character) 2.
The term container is used to refer to any entity that is capable of storing data. This entity can be a single disk device, or a group of disk devices, such as a RAIDset, stripeset, or mirrorset. The following list describes the major steps for configuring HSZ40 controller containers for use by the DEC OSF/1 operating system. 1. Add disk devices (containers) to the HSZ40 array controller configuration 2.
However, the host side SCSI host adapter’s SCSI ID is also on this bus. The HSZ40 array controller’s host side SCSI IDs must be different than the host adapter’s ID. Usually the host adapter’s ID is 7. However, this ID is usually configured through the host system console utilities and it can be something other than 7. It is recommended that the host adapter use ID 7. The HSZ40 array controller’s host side IDs are set using the HSZ40 array controller’s Command Language Interpreter (CLI).
• LUN and Logical LUN—Again, the term LUN refers to two different entities in a HSZ40 controller environment, the host side bus LUN and the device side bus LUN. Host Side LUN—This entity is part of the HSZ40 UNIT Designator. It is used by the host operating system to identify a unique HSZ40 unit. Device Side LUN—In current HSOF versions, this LUN is always 0.
Example 8–8 (Cont.) SCSI Bus Number Determination NOTICE: An external SCSI Host Adapter named PZA0 resides at PCI slot 7. It is attached to SCSI bus number 2 (SCSI--2). A disk with a name of rz18 is accessed over this bus through an HSZ40 controller.
In order for the users of this system to access the HSZ40 units associated with the DEC OSF/1 device names, the device special files associated with the DEC OSF/1 device name must have the correct major and minor numbers. The major number for units accessed through an HSZ40 controller must be ‘‘8.’’ This number is used by the Common Access Method (CAM) device driver. The minor number identifies the individual unit accessed by the CAM device driver.
The first step in creating a disk unit on the HSZ40 controller is to configure a disk device. This is done by giving the disk device a name. You must associate that name with SCSI information that allows the HSZ40 controller to relate the name to a ‘‘device bus’’ and a ‘‘target’’ on that bus. In the CLI documentation you may see the disk ‘‘name’’ defined as a ‘‘container-name’’ or a disk-device-name. To configure a disk device, you must add the disk to the controller’s configuration.
8.9 TRANSPORTABLE versus NOTRANSPORTABLE When a single device container is created it can be defined as TRANSPORTABLE or NOTRANSPORTABLE, the default is NOTRANSPORTABLE. These qualifiers are specified with the ADD DISK command. HSZ> ADD DISK DISK100 1 0 0 TRANSPORTABLE or NOTRANSPORTABLE A NOTRANSPORTABLE disk uses some of the disk space for metadata that is used by the HSZ40 array controller. A device that is specified as NOTRANSPORTABLE has a few less blocks available for user data.
To view an individual container, enter the following commands: HSZ> SHOW R0 HSZ> SHOW S0 HSZ> SHOW M0 8.11 Initialization of Controllers After containers have been created they must be initialized before units can be created.
Example 8–9 Controller Host Side SCSI IDs of 0, 1, 2, and 3 HSZ> HSZ> HSZ> HSZ> HSZ> ADD ADD ADD ADD ADD UNIT UNIT UNIT UNIT UNIT D0 DISK100 D000 DISK100 D100 R0 D200 S0 D307 DISK650 (for (for (for (for (for single device) single device) RAIDset) Stripeset) single device) The HSZ40 unit designator is used to calculate the DEC OSF/1 minor number that is used in creating the DEC OSF/1 Device Special File for an HSZ40 unit.
The default naming format that is used is rzxxx for Block Mode Device Special files and rrzxxx for Character Mode Device Special files, where the xxx denotes a number. DEC OSF/1 Device Special files as used in an HSZ40 controller subsystem, associate a disk device name with an HSZ40 unit. 8.13.1 Creating a Device Special File The following information is required to create the Device Special files for an HSZ40 unit. 1. SCSI bus number of the host side SCSI bus. 2.
Example 8–11 (Cont.) Mknod Examples OSF/1> cd /dev OSF/1> /usr/sbin/mknod rza18a b 8 34816 This creates the block mode device special file in the /dev directory for the "a" partition for this HSZ40 unit. The device special file would have the name rza18a. OSF/1> cd /dev OSF/1> /usr/sbin/mknod rrza18a c 8 34816 This creates the character mode device special file in the /dev directory for the a partition for this HSZ40 unit. The device special file would have the name rrza18a.
The xxx number determines the SCSI host side bus number and the SCSI target ID. The LUN is assumed to be 0 in all cases. The calculations are as follows: SCSI host side bus number = Integer of ( xxx / 8 ) The remainder of the above division, if any, is the target ID.
The point of these examples is to show that arbitrary DEC OSF/1 device names cannot be chosen if MAKEDEV is used to create the special files. The minor number that MAKEDEV will calculate must identify the correct HSZ40 Unit on the correct host side SCSI bus. 8.15 Helpful Utilities The DEC OSF/1 file utility can be used to determine if an HSZ40 unit can be accessed from the DEC OSF/1 host operating system. For this example, the HSZ40 has a unit designator of D101.
The device activity indicator (green light) will illuminate on the device. If the unit is a multidevice container, only one of the devices from that container illuminates.
The SCU command, scan edt, causes a polling for all devices on the host side SCSI buses. This allows you to show what devices are available from all host side SCSI buses. The special files do not have to exist for SCU to see the devices.
There should be eight entries for this device. Only the drive number will change for each entry. As an example, the configuration file for the HSZ40 unit designators D000 to D007 on host side SCSI bus # 2 are constructed as follows: The name entry is rznn. The nn is calculated as: ( 8 * Host Side SCSI Bus # ) + ( Target ID) = (8 * 2) + 0 = 16.
8.18 Using iostat and Other Utilities Some utilities do not recognize nonnumeric naming formats. Also, some utilities do not recognize a device if the device name has more than two numbers after the rz, or the utility may truncate the numbers. For example: rz445 might be seen as rz44. Some utilities cannot differentiate between rz446 or rz448. Also some utilities do not recognize the rzxnny format that has been described in this chapter, such as rza44a. One such utility is iostat.
What happens if all eight LUNs are not configured as units? ***************************************************************************** # iostat rz16 5 4 tty tin tout 0 0 0 0 0 0 0 0 rz3 tps 0 0 0 0 bps 0 0 0 0 rz4 tps 0 0 0 0 rz16 bps tps 0 0 0 0 0 0 0 0 rz16 bps tps 126 3 1618 34 1639 34 1610 34 us 1 1 1 1 cpu ni sy 0 2 0 2 0 2 0 2 id 97 98 98 98 ***************************************************************************** If all eight LUNs for a Target ID are not configured on an HSZ40, then t
The result is a kernel that contains the specific HSZ40 units that relate to the desired naming format.
A Option Order Numbers This appendix contains order numbers for controller options and preconfigured subsystems. Table A–1 lists the HSJ30 array controller options. Table A–1 HSJ30 Array Controller Options Order Number Description HSJ30–AA StorageWorks Array Cntrl 130c; 18 SCSI–2 Disk/Tape/Optical Device support; RAID 0, Base Firmware. HSJ30–AD StorageWorks Array Cntrl 130c; 18 SCSI–2 Disk/Tape/Optical Device support; 16 MB Read Cache; RAID 0, Base Firmware.
Table A–3 lists the HSJ42 subsystem options. Table A–3 HSJ42 Array Controller Subsystem Options Order Number Description HSJ42–AD Two StorageWorks Array Cntrl 140c; 36 SCSI–2 Disk/Tape/SSD /Optical Device support; 32 MB total Read Cache; RAID 0, Base Firmware. HSJ42–AF Two StorageWorks Array Cntrl 140c; 36 SCSI–2 Disk/Tape/SSD /Optical Device support; 64 MB total Read Cache; RAID 0, Base Firmware. Table A–4 lists the HSJ44 subsystem options.
Table A–6 lists the HSZ40–Ax controller options. Table A–6 HSZ40–Ax Controller Options Order Number Description HSZ40–AA StorageWorks Array Cntrl 140z; 42 SCSI–2 Disk Device support; No Cache; RAID 0, Base Firmware. HSZ40–AD StorageWorks Array Cntrl 140z; 42 SCSI–2 Disk Device support; 16 MB Read Cache; RAID 0, Base Firmware. HSZ40–AF StorageWorks Array Cntrl 140z; 42 SCSI–2 Disk Device support; 32 MB Read Cache; RAID 0, Base Firmware. HSZ40–XD 16 MB read cache module option for HSZ40–AA.
The SWKS MSC PCRM kits contain the following: • One HS operating firmware array controller software product description (SPD) (controller-specific). • One HS array controller operating firmware V2.5 release notes (controllerspecific). Refer to the related documents list in the preface of this manual for order numbers. • One StorageWorks Array Controllers HS Family of Array Controllers User’s Guide (Revision D01).
Table A–8 HSJ40 Array Controller Preconfigured Subsystems Options Order Number Description SW810–AA StorageWorks Array 140c, 6.3 GB with 6 RZ26–VA1 disk drives in an SW800–AA cabinet; 60 HZ, 120/208v 3-phase. Requires separately ordered HSJ40 array controller. SW810–AB StorageWorks Array 140c, 6.3 GB with 6 RZ26–VA1 disk drives in an SW800–AB cabinet; 50 HZ, 220/416v 3-phase. Requires separately ordered HSJ40 array controller. SW811–AA StorageWorks Array 140c, 21.
B Command Line Interpreter This appendix provides the following information: • A comprehensive list of all CLI commands • CLI error messages the operator may encounter • Examples of some common CLI-based procedures An overview of using the CLI, as well as a description of how to access and exit the CLI, is provided in Chapter 5. B.1 CLI Commands The following sections describe each of the valid commands in the CLI, along with their required parameters and qualifiers.
ADD CDROM ADD CDROM Adds a CDROM drive to the list of known CDROM drives. Format ADD CDROM container-name SCSI-location Parameters container-name Specifies the name that is used to refer to this CDROM drive. This name is referred to when creating units. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD DISK ADD DISK Adds a disk drive to the list of known disk drives. Format ADD DISK container-name SCSI-location Parameters container-name Specifies the name that is used to refer to this disk drive. This name is referred to when creating units and stripesets. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD DISK Examples 1. CLI> ADD DISK RZ26_100 1 0 0 Adds a non transportable disk to port 1, target 0, LUN 0 and names it RZ26_ 100. 2. CLI> ADD DISK DISK0 2 3 0 NOTRANSPORTABLE Adds a non transportable disk to port 2, target 3, LUN 0 and names it DISK0. 3. CLI> ADD DISK TDISK0 3 2 0 TRANSPORTABLE Adds a transportable disk to port 3, target 2, LUN 0 and names it TDISK0.
ADD LOADER ADD LOADER Note This command is valid for HSJ and HSD controllers only. Adds a loader to the list of known loaders. Format ADD LOADER container-name SCSI-location Parameters container-name Specifies the name that is used to refer to this loader. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD MIRRORSET ADD MIRRORSET Binds a set of physical devices to a mirrorset specified by a container name. Format ADD MIRRORSET container-name disk-device-name1 [disk-device-nameN] Parameters container-name Specifies the name that is used to refer to this mirrorset container. This name is referred to when creating mirrorsets. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.
ADD MIRRORSET BEST_PERFORMANCE (default) gives highest priority to finding a replacement device within the spareset that results in the best performance of the mirrorset (the device should be on a different port). If more than one device in the spareset has the best performance, the device that most closely matches the size of the remaining members of the mirrorset is selected. NOPOLICY retries a failing device from the mirrorset without selecting a replacement.
ADD OPTICAL ADD OPTICAL Adds an optical drive to the list of known optical drives. Format ADD OPTICAL container-name SCSI-location Parameters container-name Specifies the name that is used to refer to this optical drive. This name is referred to when creating units and stripesets. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD OPTICAL Examples 1. CLI> ADD OPTICAL OPT0 2 3 0 NOTRANSPORTABLE Adds a non transportable optical drive to port 2, target 3, LUN 0 and names it OPT0. 2. CLI> ADD OPTICAL TOPT0 3 2 0 TRANSPORTABLE Adds a transportable optical drive to port 3, target 2, LUN 0 and names it TOPT0.
ADD PASSTHROUGH ADD PASSTHROUGH Note This command is valid for HSJ and HSD controllers only. Creates a command disk (passthrough) container to allow direct access to a device. Format ADD PASSTHROUGH container-name device-name Parameters container-name Specifies the name that is used to refer to this passthrough container. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.
ADD RAIDSET ADD RAIDSET Creates a RAIDset from a number of containers. Format ADD RAIDSET container-name container-name1 container-name2 [container-nameN] Parameters container-name Specifies the name that is used to refer to this RAIDset. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD RAIDSET RECONSTRUCT=FAST reconstructs the RAIDset at the fastest rate possible resulting in some loss of performance of the controller overall. REDUCED NOREDUCED (Default) REDUCED specifies that the RAIDset being added is already missing one member. Use the REDUCED keyword when moving an already reduced RAIDset from one controller to another. NOREDUCED (default) identifies that all RAIDset members that make up the RAIDset are being specified. Examples 1.
ADD SPARESET ADD SPARESET Adds a disk drive to the spareset. Format ADD SPARESET disk-container-name0 [disk-container-nameN] Parameters disk-container-name0 disk-container-nameN The disk drive container names to add to the spareset. Any number of disks may be added to the spareset using only one command. Description The SPARESET is a pool of drives available to the controller to replace failing members of RAIDsets and mirrorsets.
ADD STRIPESET ADD STRIPESET Creates a stripeset from a number of containers. Format ADD STRIPESET container-name container-name1 container-name2 [container-nameN] Parameters container-name Specifies the name that is used to refer to this stripeset. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters.
ADD TAPE ADD TAPE Note This command is valid for HSJ and HSD controllers only. Adds a tape drive to the list of known tape drives. Format ADD TAPE device-name SCSI-location Parameters device-name Specifies the name that is used to refer to this tape drive. This name is referred to when creating units. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.
ADD UNIT ADD UNIT Adds a logical unit to the controller. Format ADD UNIT unit-number container-name Parameters unit-number (HSJ and HSD only) The device type letter followed by the logical unit number (0–4094) that the host uses to access the unit. The device type letter is either ‘‘D’’ for disk devices (including CDROMs) or ‘‘T’’ for tape devices.
ADD UNIT PREFERRED_PATH=THIS_CONTROLLER PREFERRED_PATH=OTHER_CONTROLLER NOPREFERRED_PATH (Default) Note The PREFERRED_PATH and NOPREFERRED_PATH qualifiers are valid for HSJ and HSD controllers only. The preferred path for HSZ controllers is determined by the unit number (the target portion of the unit number–the hundreds place) specified on the ADD UNIT command.
ADD UNIT Note The PREFERRED_PATH and NOPREFERRED_PATH qualifiers are valid for HSJ and HSD controllers only. The preferred path for HSZ controllers is determined by the unit number (the target portion of the unit number–the hundreds place) specified on the ADD UNIT command. Specifies the preferred controller that the unit should be accessed through (PREFERRED_PATH=) or whether the unit may be accessed through either controller (NOPREFERRED_PATH).
ADD UNIT Note The PREFERRED_PATH and NOPREFERRED_PATH qualifiers are valid for HSJ and HSD controllers only. The preferred path for HSZ controllers is determined by the unit number (the target portion of the unit number–the hundreds place) specified on the ADD UNIT command. Specifies the preferred controller that the unit should be accessed through (PREFERRED_PATH=) or whether the unit may be accessed through either controller (NOPREFERRED_PATH).
ADD UNIT Note When initially added, NOWRITEBACK_CACHE is the default. Qualifiers for a Unit Created from a TRANSPORTABLE Optical Drive MAXIMUM_CACHED_TRANSFER=n MAXIMUM_CACHED_TRANSFER=32 (Default) Specifies the maximum size transfer in blocks to be cached by the controller. Any transfers over this size are not cached. Valid values are 1–1024.
ADD UNIT WRITE_PROTECT NOWRITE_PROTECT (Default) Enables and disables write protection of the unit. Qualifiers for a Unit Created from a NOTRANSPORTABLE Optical Drive MAXIMUM_CACHED_TRANSFER=n MAXIMUM_CACHED_TRANSFER=32 (Default) Specifies the maximum size transfer in blocks to be cached by the controller. Any transfers over this size are not cached. Valid values are 1–1024.
ADD UNIT WRITE_PROTECT NOWRITE_PROTECT (Default) Enables and disables write protection of the unit. WRITEBACK_CACHE NOWRITEBACK_CACHE (Default) Enables and disables the controller’s write-back cache on this unit. Note It may take up to 5 minutes to flush unwritten data from the write-back cache once you disable write-back caching. Note When initially added, NOWRITEBACK_CACHE is the default.
ADD UNIT second controller will inherit any PREFERRED_PATH settings, and the two controllers will operate using the preset PREFERRED_PATH options. RUN (Default) NORUN Enables and disables a unit’s availibility to the host. When RUN (default) is specified, the devices that make up the unit will be spun up and the unit will be made available to the host. If NORUN is specified, the devices that make up the unit will still be spun up, but the unit will not be made available to the host.
ADD UNIT Specifies the preferred controller that the unit should be accessed through (PREFERRED_PATH=) or whether the unit may be accessed through either controller (NOPREFERRED_PATH). The preferred path qualifier is used only if both controllers are running in a dual-redundant configuration. If one controller fails, then all the devices will be made accessible through the remaining controller, ignoring the preferred path setting.
ADD UNIT Supported tape formats are as follows: • DEVICE_DEFAULT (default) The default tape format is the default that the device uses, or, in the case of devices that can be set via switches on the front panel, the settings of those switches.
ADD UNIT Examples 1. CLI> ADD UNIT D0 DISK0 Creates disk unit number 0 from container DISK0. 2. CLI> ADD UNIT T0 TAPE12 Creates tape unit number 0 from container TAPE12. 3. CLI> ADD UNIT D170 RAID9 WRITE_PROTECT Creates disk unit number 170 from container RAID9 and write protects it.
CLEAR_ERRORS CLI CLEAR_ERRORS CLI Stops displaying errors at the CLI prompt. Format CLEAR_ERRORS CLI Description Errors detected by controller firmware are displayed before the CLI prompt. These errors are displayed even after the error condition is rectified, until the controller is restarted or the CLEAR_ERRORS CLI command is issued. Note This command does not clear the error conditions, it only clears displaying the errors at the CLI prompt. Examples 1.
CLEAR_ERRORS INVALID_CACHE CLEAR_ERRORS INVALID_CACHE Clears all data from the cache and makes it usable by the specified controller. Format CLEAR_ERRORS INVALID_CACHE controller Parameters controller Specifies which controller will clear the INVALID_CACHE condition. Either THIS_CONTROLLER or OTHER_CONTROLLER must be specified. Description CAUTION This command causes loss of customer data.
CLEAR_ERRORS LOST_DATA CLEAR_ERRORS LOST_DATA Clears the lost data error on a unit. Format CLEAR_ERRORS LOST_DATA unit-number Parameters unit-number Specifies the logical unit number (for HSDs and HSJs D0–D4094 or T0–T4094, for HSZs D0–D7, D100-D107, and so forth) that will have the lost data error cleared. The unit-number is the name given the unit when it was created using the ADD UNIT command. Description CAUTION This command causes loss of customer data.
CLEAR_ERRORS UNKNOWN CLEAR_ERRORS UNKNOWN Clears the UNKNOWN error from a device. Format CLEAR_ERRORS UNKNOWN device-name Parameters device-name Specifies the device name of the device with the UNKNOWN error. Description Note ‘‘UNKNOWN’’ must be completely spelled out, not abbreviated. If a device has a failure such that the controller marks the device as UNKNOWN, the device is never automatically checked again to see if it has been repaired or if the failure condition was rectified.
CLEAR_ERRORS UNWRITEABLE_DATA CLEAR_ERRORS UNWRITEABLE_DATA Clears the unwriteable data error on a unit. Format CLEAR_ERRORS UNWRITEABLE_DATA unit-number Parameters unit-number Specifies the logical unit number (for HSDs and HSJs D0–D4094 or T0–T4094, for HSZs D0–D7, D100-D107, and so forth) that will have the unwriteable data error cleared. The unit-number is the name given the unit when it was created using the ADD UNIT command. Description CAUTION This command causes loss of customer data.
DELETE container-name DELETE container-name Deletes a container from the list of known containers. Format DELETE container-name Parameters container-name Specifies the name that identifies the container. This is the name given the container when it was created using the ADD command (ADD DEVICE, ADD STRIPESET, and so forth). Description Checks to see if the container is used by any other containers or a unit. If the container is in use, an error is displayed and the container is not deleted.
DELETE FAILEDSET DELETE FAILEDSET Delete a disk drive from the failedset. Format DELETE FAILEDSET disk-container-name0 [disk-container-nameN] Parameters disk-container-name0 disk-container-nameN The disk drive container names to delete from the failedset. Any number of disks may be deleted from the failedset using only one command.
DELETE SPARESET DELETE SPARESET Delete a disk drive from the spareset. Format DELETE SPARESET disk-container-name0 [disk-container-nameN] Parameters disk-container-name0 disk-container-nameN The disk drive container names to delete from the spareset. Any number of disks may be deleted from the spareset using only one command. Description The SPARESET is a pool of drives available to the controller to replace failing members of RAIDsets and mirrorsets.
DELETE unit-number DELETE unit-number Deletes a unit from the list of known units. Format DELETE unit-number Parameters unit-number Specifies the logical unit number (for HSDs and HSJs D0–D4094 or T0–T4094, for HSZs D0–D7, D100-D107, and so forth) that is to be deleted. The unit-number is the name given the unit when it was created using the ADD UNIT command. Description The DELETE command flushes any user data from the write-back cache to the disk and deletes the logical unit.
DIRECTORY DIRECTORY Lists the diagnostics and utilities available on THIS_CONTROLLER. Format DIRECTORY Description The DIRECTORY command lists the various diagnostics and utilities that are available on THIS_CONTROLLER. A directory of diagnostics and utilities available on this controller is displayed. For specific information about the diagnostics and utilities available, refer to the StorageWorks Array Controllers HS Family of Array Controllers Service Manual. Examples 1.
EXIT EXIT Exits the CLI and breaks the virtual terminal connection. Format EXIT Description When entering the EXIT command from a host using a virtual terminal connection, the connection is broken and control is returned to the host. If entered from a maintenance terminal, the EXIT command restarts the CLI, displaying the copyright notice, the controller type, and the last fail packet. Examples 1.
HELP HELP Displays an overview for getting help. Format HELP Description The HELP command displays a brief description for using the question mark ‘‘?’’ to obtain help on any command or CLI function. Examples 1. CLI> HELP Help may be requested by typing a question mark (?) at the CLI prompt. This will print a list of all available commands For further information you may enter a partial command and type a space followed by a "?" to print a list of all available options at that point in the command.
INITIALIZE INITIALIZE Initializes the metadata on the container specified. Format INITIALIZE container-name Parameters container-name Specifies the container name to initialize. Description The INITIALIZE command initializes a container so a logical unit may be created from it. During initialization, a small amount of disk space is used for controller metadata and is made inaccessible to the host.
INITIALIZE DESTROY (Default) NODESTROY This qualifier prevents the user data and forced error metadata from being destroyed during intialization. This allows the data on the container to be reused for a disk, stripeset, or mirrorset unit. (The NODESTROY qualifier is ignored for RAIDsets.) NODESTROY is only used when creating a unit out of devices that have been reduced from mirrorsets. Examples 1. CLI> INITIALIZE DISK0 Initializes container DISK0.
LOCATE LOCATE Locates units, storagesets, and devices by lighting the amber device fault LED on the front of the StorageWorks building block (SBB). Format LOCATE Description The LOCATE command illuminates the amber device fault LEDs (the lower LED on the front of an SBB) of the containers specified. The LOCATE command also can be used as a lamp test. Qualifiers ALL The LOCATE ALL command turns on the amber device fault LEDs of all configured devices. This qualifier also can be used as a lamp test.
LOCATE UNITS The LOCATE UNITS command turns on the amber device fault LEDs of all devices used by units. This command is useful to determine which devices are not currently configured into logical units. See LOCATE CANCEL to turn off device the LEDs. An error is displayed if no units have been configured. PTL SCSI-location The LOCATE PTL SCSI-location command turns on the amber device fault LEDs at the given SCSI location.
MIRROR disk-device-name1 container-name MIRROR disk-device-name1 container-name Allows you to convert a physical device to a one-member mirrorset. Format MIRROR disk-device-name1 container-name Parameters disk-device-name1 Specifies the name of the physical device that you wish to convert to a onemember mirrorset. The device must be part of a unit. container-name Specifies the name that is used to refer to this mirrorset.
MIRROR disk-device-name1 container-name POLICY=BEST_FIT POLICY=BEST_PERFORMANCE NOPOLICY (Default) The POLICY qualifier specifies the replacement policy to be used when a mirrorset member within the mirrorset fails. BEST_FIT gives highest priority to finding a replacement device within the spareset that most closely matches the sizes of the remaining members of the mirrorset. If more than one device in the spareset is the correct size, the device that gives the best performance is selected.
REDUCE disk-device-name1 [disk-device-nameN] REDUCE disk-device-name1 [disk-device-nameN] Allows you to remove members from an existing mirrorset. Format REDUCE disk-device-name1 [disk-device-nameN] Parameters disk-device-name1 Specifies the name of the NORMAL mirrorset member to be removed. [disk-device-nameN] Specifies the name of the second mirrorset member to be removed. Description This command allows you to remove members from mirrorsets.
REDUCE disk-device-name1 [disk-device-nameN] Examples 1. CLI> REDUCE DISK210 DISK110 DISK210 and DISK110 are removed from their respective mirrorsets.
RENAME RENAME Renames a container. Format RENAME old-container-name new-container-name Parameters old-container-name Specifies the existing name that identifies the container. new-container-name Specifies the new name to identify the container. This name is referred to when creating units and storagesets. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.
RESTART OTHER_CONTROLLER RESTART OTHER_CONTROLLER Restarts the other controller. Format RESTART OTHER_CONTROLLER Description The RESTART OTHER_CONTROLLER command flushes all user data from the other controller’s write-back cache (if present), then restarts the other controller. If any disks are online to the other controller, the controller does not restart unless the OVERRIDE_ONLINE qualifier is specified (HSD and HSJ only).
RESTART OTHER_CONTROLLER OVERRIDE_ONLINE NOOVERRIDE_ONLINE (Default) If any units are online to the controller, the controller is not restarted unless OVERRIDE_ONLINE is specified. If the OVERRIDE_ONLINE qualifier is specified, the controller restarts after all customer data is written to disk. CAUTION Customer data may be lost or corrupted if the OVERRIDE_ONLINE qualifier is specified.
RESTART THIS_CONTROLLER RESTART THIS_CONTROLLER Restarts this controller. Format RESTART THIS_CONTROLLER Description The RESTART THIS_CONTROLLER command flushes all user data from this controller’s write-back cache (if present), then restarts this controller. If any disks are online to the other controller, the controller does not restart unless the OVERRIDE_ONLINE qualifier is specified (HSD and HSJ only).
RESTART THIS_CONTROLLER CAUTION Customer data may be lost or corrupted if the IMMEDIATE qualifier is specified. OVERRIDE_ONLINE NOOVERRIDE_ONLINE (Default) If any units are online to the controller, the controller is not restarted unless OVERRIDE_ONLINE is specified. If the OVERRIDE_ONLINE qualifier is specified, the controller restarts after all customer data is written to disk. CAUTION Customer data may be lost or corrupted if the OVERRIDE_ONLINE qualifier is specified.
RESTART THIS_CONTROLLER 2. CLI> RESTART THIS_CONTROLLER OVERRIDE_ONLINE Restarts this controller even if there are units online to this controller.
RETRY_ERRORS UNWRITEABLE_DATA RETRY_ERRORS UNWRITEABLE_DATA Tries to write the unwriteable data on a unit. Format RETRY_ERRORS UNWRITEABLE_DATA unit-number Parameters unit-number Specifies the logical unit number (for HSDs and HSJs D0–D4094 or T0–T4094, for HSZs D0–D7, D100-D107, and so forth) which the write operation of the unwriteable data is attempted. The unit-number is the name given the unit when it was created using the ADD UNIT command.
RUN RUN Runs a diagnostic or utility on THIS_CONTROLLER. Format RUN program-name Parameters program-name The name of the diagnostic or utility to be run. DILX and CLONE are examples of utilities and diagnostics that can be run from the CLI. Description The RUN command starts various diagnostics and utilities on THIS_CONTROLLER. Diagnostics and utilities can be run only on the controller where the terminal or DUP connection is connected.
SELFTEST OTHER_CONTROLLER SELFTEST OTHER_CONTROLLER Runs a self-test on the other controller. Format SELFTEST OTHER_CONTROLLER Description The SELFTEST OTHER_CONTROLLER command flushes all user data from the other controller’s write-back cache (if present), shuts down the other controller, then restarts it in DAEMON loop-on-self-test mode. The OCP reset (//) button must be pressed to take the other controller out of loop-on-self-test mode.
SELFTEST OTHER_CONTROLLER If the OVERRIDE_ONLINE qualifier is specified, the controller starts the self-test after all customer data is written to disk from the write-back cache. CAUTION Customer data may be lost or corrupted if the OVERRIDE_ONLINE qualifier is specified. Qualifiers for HSZ Controllers IGNORE_ERRORS NOIGNORE_ERRORS (Default) If errors result when trying to write user data, the controller does not start the self-test unless IGNORE_ERRORS is specified.
SELFTEST THIS_CONTROLLER SELFTEST THIS_CONTROLLER Runs a self-test on this controller. Format SELFTEST THIS_CONTROLLER Description The SELFTEST THIS_CONTROLLER command flushes all user data from this controller’s write-back cache (if present), shuts down this controller, then restarts it in DAEMON loop-on-self-test mode. The OCP reset (//) button must be pressed to take this controller out of loop-on-self-test mode.
SELFTEST THIS_CONTROLLER CAUTION Customer data may be lost or corrupted if the IMMEDIATE qualifier is specified. OVERRIDE_ONLINE NOOVERRIDE_ONLINE (Default) If any units are online to the controller, the controller does not self-test unless OVERRIDE_ONLINE is specified. If the OVERRIDE_ONLINE qualifier is specified, the controller starts the self-test after all customer data is written to disk from the write-back cache.
SET disk-container-name SET disk-container-name Changes the transportable characteristics of a disk drive. Format SET disk-container-name Parameters disk-container-name The name of the disk drive that will have its characteristics changed. Description Changes the characteristics of a disk drive.
SET FAILOVER SET FAILOVER Places THIS_CONTROLLER and OTHER_CONTROLLER into a dual-redundant configuration. Format SET FAILOVER COPY=configuration-source Parameters COPY=configuration-source Specifies where the ‘‘good’’ copy of the device configuration resides.
SET FAILOVER Examples 1. CLI> SET FAILOVER COPY=THIS_CONTROLLER Places two controllers into a dual-redundant configuration, where the ‘‘good’’ data was on the controller that the maintenance terminal or virtual terminal connection was connected to. 2. CLI> SET FAILOVER COPY=OTHER_CONTROLLER Places two controllers into a dual-redundant configuration, where the ‘‘good’’ data was on the controller that the maintenance terminal or virtual terminal connection was not connected to.
SET mirrorset-container-name SET mirrorset-container-name Changes the characteristics of a mirrorset. Format SET mirrorset-container-name Parameters mirrorset-container-name The name of the mirrorset that will have its characteristics modified. The name must start with a letter (A–Z) and can then consist of up to eight more characters made up of letters A–Z, numbers 0–9, periods (.), dashes (-), or underscores (_), for a total of nine characters. Description Changes the characteristics of a mirrorset.
SET mirrorset-container-name REMOVE=disk-container-name The SET mirrorset-container-name REMOVE=disk-device-name CLI command allows you to remove members from an existing mirrorset. The device specified by disk-device-name is removed from the mirrorset specified by mirrorset-containername. If the physical device is not a member of the mirrorset, or if the mirrorset will not have a remaining NORMAL or NORMALIZING member, then an error is reported and no action is taken.
SET mirrorset-container-name NOPOLICY retries a failing device from the mirrorset without selecting a replacement. This causes the mirrorset to run with less than the nominal number of members until a BEST_FIT or BEST_PERFORMANCE policy is selected, or a member is manually replaced in the mirrorset. READ_SOURCE=read-source The SET mirrorset-container-name READ_SOURCE=read-source command allows you to control the read algorithm for the specified mirrorset.
SET NOFAILOVER SET NOFAILOVER Removes THIS_CONTROLLER and OTHER_CONTROLLER (if reachable) from a dual-redundant configuration. Format SET NOFAILOVER Description The SET NOFAILOVER command removes THIS_CONTROLLER and the OTHER_CONTROLLER (if currently reachable) from a dual-redundant configuration. Before or immediately after entering this command, one controller should be physically removed because the sharing of devices is not supported by single controller configurations.
SET OTHER_CONTROLLER SET OTHER_CONTROLLER Changes the other controller’s parameters (in a dual-redundant configuration, the controller that the maintenance terminal is not connected to, or the controller that is not the target of the DUP connection. Format SET OTHER_CONTROLLER Description The SET OTHER_CONTROLLER command allows you to modify the controller parameters of the other controller in a dual-redundant configuration.
SET OTHER_CONTROLLER If the batteries go low after controller initialization, unwritten cache data is flushed from the cache and any RAIDset or mirrorset that does not have access to good batteries is made inoperative, regardless of the cache policy. • If the batteries are bad or missing, RAIDsets and mirrorsets are made inoperative. • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced.
SET OTHER_CONTROLLER When first installed, the controller’s terminal speed is set to 9600 baud. TMSCP_ALLOCATION_CLASS=n Specifies the allocation class (0–255 in a single controller configuration or 1–255 in a dual-redundant configuration). When first installed, the controller’s TMSCP_ALLOCATION_CLASS is set to 0. Qualifiers for HSJ Controllers CI_ARBITRATION=ASYNCHRONOUS (Default) CI_ARBITRATION=SYNCHRONOUS ASYNCHRONOUS arbitration is currently (Version 2.5) implemented in HSJ HSOF.
SET OTHER_CONTROLLER • If the batteries are bad or missing, RAIDsets and mirrorsets are made inoperative. • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced. CAUTION There is some risk in setting CACHE_POLICY=B to allow access to mirrorsets and RAIDsets when the batteries are low. Because the batteries may be in an unknown state, there is no guarantee as to how long they will maintain data in the cache if a power failure occurs.
SET OTHER_CONTROLLER When first installed, the controller’s terminal parity is set to NOTERMINAL_PARITY. TERMINAL_SPEED=baud_rate Sets the terminal speed to 300, 600, 1200, 2400, 4800, 9600 or 19200 baud. The transmit speed is always equal to the receive speed. When first installed, the controller’s terminal speed is set to 9600 baud. TMSCP_ALLOCATION_CLASS=n Specifies the allocation class (0–255 in a single controller configuration or 1–255 in a dual-redundant configuration).
SET OTHER_CONTROLLER • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced. CAUTION There is some risk in setting CACHE_POLICY=B to allow access to mirrorsets and RAIDsets when the batteries are low. Because the batteries may be in an unknown state, there is no guarantee as to how long they will maintain data in the cache if a power failure occurs. Fully charged batteries will maintain the data for a minimum of 100 hours.
SET OTHER_CONTROLLER TERMINAL_PARITY=ODD TERMINAL_PARITY=EVEN NOTERMINAL_PARITY Specifies the parity transmitted and expected. Parity options are ODD or EVEN. NOTERMINAL_PARITY causes the controller to not check for, or transmit any parity on the terminal lines. When first installed, the controller’s terminal parity is set to NOTERMINAL_PARITY. TERMINAL_SPEED=baud_rate Sets the terminal speed to 300, 600, 1200, 2400, 4800, 9600 or 19200 baud. The transmit speed is always equal to the receive speed.
SET RAIDset-container-name SET RAIDset-container-name Changes the characteristics of a RAIDset. Format SET RAIDset-container-name Parameters RAIDset-container-name The name of the RAIDset that will have its characteristics modified. Description Changes the characteristics of a RAIDset. Qualifiers POLICY=BEST_FIT POLICY=BEST_PERFORMANCE (Default) NOPOLICY Specifies the replacement policy to use when a member within the RAIDset fails.
SET RAIDset-container-name specified, the RAIDset continues to operate in a reduced state until a replacement is manually specified (see SET RAIDset-container-name) REPLACE=) or a policy is specified (see SET RAIDset-container-name POLICY=). The disk removed via the REMOVE= command is added to the failedset. Note No other qualifiers to the SET RAIDset-container-name command may be specified if REMOVE is specified.
SET THIS_CONTROLLER SET THIS_CONTROLLER Changes this controller’s parameters (the controller that the maintenance terminal is connected to or the target of the DUP connection). Format SET THIS_CONTROLLER Description The SET THIS_CONTROLLER command allows you to modify controller parameters on THIS_CONTROLLER in single and dual-redundant configurations.
SET THIS_CONTROLLER • If the batteries are bad or missing, RAIDsets and mirrorsets are made inoperative. • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced. CAUTION There is some risk in setting CACHE_POLICY=B to allow access to mirrorsets and RAIDsets when the batteries are low. Because the batteries may be in an unknown state, there is no guarantee as to how long they will maintain data in the cache if a power failure occurs.
SET THIS_CONTROLLER TMSCP_ALLOCATION_CLASS=n Specifies the allocation class (0–255 in a single controller configuration or 1–255 in a dual-redundant configuration). When first installed, the controller’s TMSCP_ALLOCATION_CLASS is set to 0. Qualifiers for HSJ Controllers CI_ARBITRATION=ASYNCHRONOUS (Default) CI_ARBITRATION=SYNCHRONOUS ASYNCHRONOUS arbitration is currently (Version 2.5) implemented in HSJ HSOF. SYNCHRONOUS is for future CI host adapters.
SET THIS_CONTROLLER • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced. CAUTION There is some risk in setting CACHE_POLICY=B to allow access to mirrorsets and RAIDsets when the batteries are low. Because the batteries may be in an unknown state, there is no guarantee as to how long they will maintain data in the cache if a power failure occurs. Fully charged batteries will maintain the data for a minimum of 100 hours.
SET THIS_CONTROLLER TERMINAL_SPEED=baud_rate Sets the terminal speed to 300, 600, 1200, 2400, 4800, 9600 or 19200 baud. The transmit speed is always equal to the receive speed. When first installed, the controller’s terminal speed is set to 9600 baud. TMSCP_ALLOCATION_CLASS=n Specifies the allocation class (0–255 in a single controller configuration or 1–255 in a dual-redundant configuration). When first installed, the controller’s TMSCP_ALLOCATION_CLASS is set to 0.
SET THIS_CONTROLLER • Write-back caching automatically resumes when the cache batteries are fully recharged or replaced. CAUTION There is some risk in setting CACHE_POLICY=B to allow access to mirrorsets and RAIDsets when the batteries are low. Because the batteries may be in an unknown state, there is no guarantee as to how long they will maintain data in the cache if a power failure occurs. Fully charged batteries will maintain the data for a minimum of 100 hours.
SET THIS_CONTROLLER TERMINAL_PARITY=ODD TERMINAL_PARITY=EVEN NOTERMINAL_PARITY Specifies the parity transmitted and expected. Parity options are ODD or EVEN. NOTERMINAL_PARITY causes the controller to not check for, or transmit any parity on the terminal lines. When first installed, the controller’s terminal parity is set to NOTERMINAL_PARITY. TERMINAL_SPEED=baud_rate Sets the terminal speed to 300, 600, 1200, 2400, 4800, 9600 or 19200 baud. The transmit speed is always equal to the receive speed.
SET unit-number SET unit-number Changes the unit parameters. Format SET unit-number Parameters unit-number Specifies the logical unit number (for HSDs and HSJs D0–D4094 or T0–T4094, for HSZs D0–D7, D100-D107, and so forth) to modify the software switches. The unit-number is the name given the unit when it was created using the ADD UNIT command. Description The SET command is used to change logical unit parameters.
SET unit-number Note The PREFERRED_PATH qualifier may be specified on a single controller, however, the qualifier will not take effect until a second controller is added and the two controllers are configured for dual-redundancy. The second controller will inherit any PREFERRED_PATH settings, and the two controllers will operate using the preset PREFERRED_PATH options. READ_CACHE (Default) NOREAD_CACHE Enables and disables the controller’s read cache on this unit.
SET unit-number second controller will inherit any PREFERRED_PATH settings, and the two controllers will operate using the preset PREFERRED_PATH options. READ_CACHE (Default) NOREAD_CACHE Enables and disables the controller’s read cache on this unit. RUN (Default) NORUN Enables and disables a unit’s availibility to the host. When RUN (default) is specified, the devices that make up the unit will be spun up and the unit will be made available to the host.
SET unit-number second controller will inherit any PREFERRED_PATH settings, and the two controllers will operate using the preset PREFERRED_PATH options. READ_CACHE (Default) NOREAD_CACHE Enables and disables the controller’s read cache on this unit. RUN (Default) NORUN Enables and disables a unit’s availibility to the host. When RUN (default) is specified, the devices that make up the unit will be spun up and the unit will be made available to the host.
SET unit-number The preferred path qualifier is used only if both controllers are running in a dual-redundant configuration. If one controller fails, then all the devices will be made accessible through the remaining controller, ignoring the preferred path setting. When the failed controller is restarted, the drives automatically return to the controller specified by the preferred path qualifier.
SET unit-number The preferred path qualifier is used only if both controllers are running in a dual-redundant configuration. If one controller fails, then all the devices will be made accessible through the remaining controller, ignoring the preferred path setting. When the failed controller is restarted, the drives automatically return to the controller specified by the preferred path qualifier.
SET unit-number Note The PREFERRED_PATH and NOPREFERRED_PATH qualifiers are valid for HSJ and HSD controllers only. The preferred path for HSZ controllers is determined by the unit number (the target portion of the unit number–the hundreds place) specified on the ADD UNIT command. Specifies the preferred controller that the unit should be accessed through (PREFERRED_PATH=) or whether the unit may be accessed through either controller (NOPREFERRED_PATH).
SET unit-number Note It may take up to 5 minutes to flush unwritten data from the write-back cache once you disable write-back caching. Qualifiers for a Unit Created from a Stripeset MAXIMUM_CACHED_TRANSFER=n MAXIMUM_CACHED_TRANSFER=32 (Default) Specifies the maximum size transfer in blocks to be cached by the controller. Any transfers over this size are not cached. Valid values are 1–1024.
SET unit-number made available to the host. If NORUN is specified, the devices that make up the unit will still be spun up, but the unit will not be made available to the host. WRITE_PROTECT NOWRITE_PROTECT (Default) Enables and disables write protection of the unit. WRITEBACK_CACHE NOWRITEBACK_CACHE (Default) Enables and disables the controller’s write-back cache on this unit. Note It may take up to 5 minutes to flush unwritten data from the write-back cache once you disable write-back caching.
SET unit-number PREFERRED_PATH=THIS_CONTROLLER PREFERRED_PATH=OTHER_CONTROLLER NOPREFERRED_PATH (Default) Note The PREFERRED_PATH and NOPREFERRED_PATH qualifiers are valid for HSJ and HSD controllers only. The preferred path for HSZ controllers is determined by the unit number (the target portion of the unit number–the hundreds place) specified on the ADD UNIT command.
SHOW CDROMS SHOW CDROMS Shows all CD–ROM drives and drive information. Format SHOW CDROMS Description The SHOW CDROMS command displays all the CD–ROM drives known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Examples 1.
SHOW cdrom-container-name SHOW cdrom-container-name Shows information about a CD–ROM. Format SHOW cdrom-container-name Parameters cdrom-container-name The name of the CD–ROM drive to be displayed. Description The SHOW cdrom-container-name command is used to show specific information about a particular CD–ROM drive. Examples 1.
SHOW DEVICES SHOW DEVICES Shows physical devices and physical device information. Format SHOW DEVICES Description The SHOW DEVICES command displays all the devices known to the controller. First disks are shown, then tapes, then CDROMs and finally opticals. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Information contained in the additional information is dependent on the device type. Examples 1.
SHOW DISKS SHOW DISKS Shows all disk drives and drive information. Format SHOW DISKS Description The SHOW DISKS command displays all the disk drives known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Examples 1.
SHOW disk-container-name SHOW disk-container-name Shows information about a disk drive. Format SHOW disk-container-name Parameters disk-container-name The name of the disk drive to be displayed. Description The SHOW disk-container-name command is used to show specific information about a particular disk. Examples 1.
SHOW FAILEDSET SHOW FAILEDSET Shows the members of the failedset. Format SHOW FAILEDSET Description The SHOW FAILEDSET command displays all the disk drives that are members of the failedset. Examples 1. CLI> SHOW FAILEDSET Name Storageset Uses Used by -----------------------------------------------------------------------------FAILEDSET failedset DISK310 DISK410 Shows a listing of the members of the failedset.
SHOW LOADERS SHOW LOADERS Note This command is valid for HSJ and HSD controllers only. Shows all loaders and loader information. Format SHOW LOADERS Description The SHOW LOADERS command displays all the loaders known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Examples 1.
SHOW loader-container-name SHOW loader-container-name Note This command is valid for HSJ and HSD controllers only. Shows information about a loader. Format SHOW loader-container-name Parameters loader-container-name The name of the loader to be displayed. Description The SHOW loader-container-name command is used to show specific information about a particular loader. Examples CLI> SHOW LDR511 1.
SHOW MIRRORSETS SHOW MIRRORSETS Shows all configured mirrorsets and any mirrorset-specific data related only to mirrorsets. Format SHOW MIRRORSETS Description The SHOW MIRRORSETS command displays all the mirrorsets known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each mirrorset. Examples 1.
SHOW MIRRORSETS MIRR3 mirrorset DISK120 DISK250 STR0 Switches: POLICY (for replacement) = BEST_PERFORMANCE COPY (priority) = FAST READ_SOURCE = LEAST_BUSY MEMBERSHIP = 2, 2 members present State: DISK250 (member 0) is NORMAL DISK120 (member 1) is NORMAL Size: 4109470 blocks MIRR4 mirrorset DISK330 Switches: POLICY (for replacement) = BEST_PERFORMANCE COPY (priority) = NORMAL READ_SOURCE = LEAST_BUSY MEMBERSHIP = 1, 1 member present State: DISK330 (member 0) is NORMAL Size: 2050353 blocks STR0 Shows
SHOW mirrorset-container-name SHOW mirrorset-container-name Shows the same information as SHOW MIRRORSETS FULL except that it only displays information on the mirrorset specified by mirrorset-container-name. Format SHOW mirrorset-container-name Parameters mirrorset-container-name The name of the mirrorset to be displayed. Description The SHOW mirrorset-container-name command is used to show specific information about a particular mirrorset. Examples 1.
SHOW OPTICALS SHOW OPTICALS Shows all optical drives and drive information. Format SHOW OPTICALS Description The SHOW OPTICALS command displays all the optical drives known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Examples 1.
SHOW optical-container-name SHOW optical-container-name Shows information about an optical drive. Format SHOW optical-container-name Parameters optical-container-name The name of the optical drive to be displayed. Description The SHOW optical-container-name command is used to show specific information about a particular optical drive. Examples 1.
SHOW OTHER_CONTROLLER SHOW OTHER_CONTROLLER Shows information for the other controller. Format SHOW OTHER_CONTROLLER Description Shows all controller, port, and terminal information for the other controller. Qualifiers FULL If the FULL qualifier is specified, additional information is displayed after the basic controller information. Examples 1. CLI> SHOW OTHER_CONTROLLER Controller: HSJ40 (C) DEC ZG01234567 Firmware V2.
SHOW OTHER_CONTROLLER MSCP allocation class TMSCP allocation class 9 9 Cache: 32 megabyte write cache, version 2 Cache is GOOD Battery is GOOD No unflushed data in cache CACHE_FLUSH_TIMER = DEFAULT (10 seconds) CACHE_POLICY = A Shows the basic HSD controller information. 3. CLI> SHOW OTHER_CONTROLLER Controller: HSZ40 (C) DEC CX44332211 Firmware V2.
SHOW PASSTHROUGH SHOW PASSTHROUGH Shows passthrough containers and container information. Format SHOW PASSTHROUGH Description The SHOW PASSTHROUGH command displays all the passthrough containers known by the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each storageset. Examples 1.
SHOW passthrough-container-name SHOW passthrough-container-name Shows information about a passthrough container. Format SHOW passthrough-container-name Parameters passthrough-container-name The name of the passthrough container to be displayed. Description The SHOW passthrough-container-name command is used to show specific information about a passthrough container. Examples 1.
SHOW RAIDSETS SHOW RAIDSETS Shows RAIDsets and RAIDset information. Format SHOW RAIDSETS Description The SHOW RAIDSETS command displays all the RAIDsets known by the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each storageset. Examples 1.
SHOW RAIDSETS R1 raidset Switches: POLICY (for replacement) RECONSTRUCT (priority) = CHUNKSIZE = 63 blocks State: NORMAL DISK130 (member 0) is DISK240 (member 1) is DISK330 (member 2) is DISK420 (member 3) is Size: 2050353 blocks DISK130 DISK240 DISK330 DISK420 = BEST_PERFORMANCE NORMAL NORMAL NORMAL NORMAL NORMAL Shows a full listing of all RAIDsets.
SHOW raidset-container-name SHOW raidset-container-name Shows information about a RAIDset. Format SHOW raidset-container-name Parameters raidset-container-name The name of the RAIDset to be displayed. Description The SHOW raidset-container-name command is used to show specific information about a particular RAIDset. Examples 1.
SHOW SPARESET SHOW SPARESET Shows the members of the spareset. Format SHOW SPARESET Description The SHOW SPARESET command displays all the disk drives that are members of the spareset. Examples 1. CLI> SHOW SPARESET Name Storageset Uses Used by -----------------------------------------------------------------------------SPARESET spareset Shows a list of the members of the spareset.
SHOW STORAGESETS SHOW STORAGESETS Shows storagesets and storageset information. Format SHOW STORAGESETS Description The SHOW STORAGESETS command displays all the storagesets known by the controller. A storageset is any collection of containers, such as stripesets, mirrorsets, RAIDsets, the spareset and the failedset. Stripesets are displayed first, followed by mirrorsets, RAIDsets, sparesets, failedsets, and then passthrough containers.
SHOW STORAGESETS S1 stripeset DISK620 DISK640 Switches: CHUNKSIZE = 24 blocks State: NORMAL DISK620 (member 0) is NORMAL DISK640 (member 1) is NORMAL Size: 31304354 blocks R0 raidset Switches: POLICY (for replacement) RECONSTRUCT (priority) = CHUNKSIZE = 63 blocks State: NORMAL DISK110 (member 0) is DISK220 (member 1) is DISK310 (member 2) is DISK400 (member 3) is Size: 2050353 blocks R1 DISK110 DISK220 DISK310 DISK400 = BEST_PERFORMANCE NORMAL NORMAL NORMAL NORMAL NORMAL raidset Switches: POLICY
SHOW STRIPESETS SHOW STRIPESETS Shows stripesets and related stripeset information. Format SHOW STRIPESETS Description The SHOW STRIPESET command displays all the stripesets known by the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each storageset. Examples 1.
SHOW STRIPESETS S1 stripeset Switches: CHUNKSIZE = 24 blocks State: NORMAL DISK620 (member 0) is NORMAL DISK640 (member 1) is NORMAL Size: 2050353 blocks Shows a full listing of all stripesets.
SHOW stripeset-container-name SHOW stripeset-container-name Shows information about a specific stripeset. Format SHOW stripeset-container-name Parameters stripeset-container-name The name of the stripeset to be displayed. Description The SHOW stripeset-container-name command is used to show specific information about a particular stripeset. Examples 1.
SHOW TAPES SHOW TAPES Note This command is valid for HSJ and HSD controllers only. Shows all tape drives and tape drive information. Format SHOW TAPES Description The SHOW TAPES command displays all the tape drives known to the controller. Qualifiers FULL If the FULL qualifier is specified, additional information may be displayed after each device. Examples 1.
SHOW tape-container-name SHOW tape-container-name Note This command is valid for HSJ and HSD controllers only. Shows information about a specific tape drive. Format SHOW tape-container-name Parameters tape-container-name The name of the tape drive to be displayed. Description The SHOW tape-container-name command is used to show specific information about a particular tape drive. Examples 1.
SHOW THIS_CONTROLLER SHOW THIS_CONTROLLER Shows information for this controller. Format SHOW THIS_CONTROLLER Description Shows all controller, port, and terminal information for this controller. Qualifiers FULL If the FULL qualifier is specified, additional information is displayed after the basic controller information. Examples 1.
SHOW THIS_CONTROLLER 2.
SHOW THIS_CONTROLLER 4.
SHOW UNITS SHOW UNITS Shows all units and unit information. Format SHOW UNITS Description The SHOW UNITS command displays all the units known by the controller. First disks (including CDROMs) are listed, then tapes. Qualifiers FULL If the FULL qualifier is specified after UNITS, additional information may be displayed after each unit-number, such as the switch settings. Examples 1.
SHOW unit-number SHOW unit-number Shows information about a specific unit. Format SHOW unit-number Parameters unit-number The unit number of the unit that is to be displayed. Description The SHOW unit-number command is used to show specific information about a particular unit. Examples 1.
SHUTDOWN OTHER_CONTROLLER SHUTDOWN OTHER_CONTROLLER Shuts down and does not restart the other controller. Format SHUTDOWN OTHER_CONTROLLER Description The SHUTDOWN OTHER_CONTROLLER command flushes all user data from the other controller’s write-back cache (if present), then shuts down the other controller. If any disks are online to the other controller, the controller does not shut down unless the OVERRIDE_ONLINE qualifier is specified (HSD and HSJ only).
SHUTDOWN OTHER_CONTROLLER If the OVERRIDE_ONLINE qualifier is specified, the controller shuts down after all customer data is written to disk. CAUTION Customer data may be lost or corrupted if the OVERRIDE_ONLINE qualifier is specified. Qualifiers for HSZ Controllers IGNORE_ERRORS NOIGNORE_ERRORS (Default) If errors result when trying to write user data, the controller is not shut down unless IGNORE_ERROR is specified.
SHUTDOWN THIS_CONTROLLER SHUTDOWN THIS_CONTROLLER Shuts down and does not restart this controller. Format SHUTDOWN THIS_CONTROLLER Description The SHUTDOWN THIS_CONTROLLER command flushes all user data from this controller’s write-back cache (if present), then shuts down this controller. If any disks are online to this controller, the controller does not shut down unless the OVERRIDE_ONLINE qualifier is specified (HSD and HSJ only).
SHUTDOWN THIS_CONTROLLER OVERRIDE_ONLINE NOOVERRIDE_ONLINE (Default) If any units are online to the controller, the controller is not shutdown unless OVERRIDE_ONLINE is specified. If the OVERRIDE_ONLINE qualifier is specified, the controller shuts down after all customer data is written to disk. CAUTION Customer data may be lost or corrupted if the OVERRIDE_ONLINE qualifier is specified.
UNMIRROR disk-device-name UNMIRROR disk-device-name Converts a one-member mirrorset back to a single device and deletes the mirrorset from the list of known mirrorsets. Format UNMIRROR disk-device-name Description Allows you to convert a mirrorset with one disk, specified by the disk name, disk-device-name, to a physical device. This command can be used on mirrorsets that are already members of higher level containers (stripesets or units). Examples 1.
B.2 CLI Messages The following sections describe messages you can encounter during interactive use of the CLI. B.2.1 Error Conventions An Error nnnn: message means that the command did not complete. Except for a few of the failover messages (6000 series), no part of the command was executed. When encountering an error entering or exiting dual-redundant mode, some synchronization problems are unavoidable; the error message in such a case tells you what to do to get things back in synchronization.
Error 1110: Unit numbers may not have leading zeros Explanation: Tape and disk unit numbers may not be of the form ‘‘D03,’’ for example, ‘‘D3’’ should be specified. Retry the ADD command without any leading zeros. Error 1120: LUN is already used Explanation: Lun number has already been used by a disk. Retry the ADD command specifying a different LUN. Error 1130: The unit number cannot exceed Explanation: You specified a unit number that was out-of-bounds.
Error 2030: This port, target LUN combination already in use by another device Explanation: When adding a device, you specified PTL that is already specified by another device. Error 2040: Cannot set TRANSPORTABLE when device in use by an upper layer Explanation: A disk cannot be set to TRANSPORTABLE when it is being used by an upper level (unit or storageset).
Error 3070: is not a member of the failedset Explanation: You attempted to delete a disk drive from the failedset that was not a member of the failedset. Error 3080: can’t be deleted Explanation: You attempted to delete the spareset or the failedset. These containers cannot be deleted. Error 3090: support is not enabled on this controller Explanation: You attempted to use a feature that requires a license, and the license was not enabled on this controller.
Error 3170: is not reduced. Cannot replace a member Explanation: When issuing a SET REPLACE=, the container specified was not reduced. Remove a member before replacing it. Error 3180: has a replacement policy specified. Cannot manually replace a member. Explanation: When issuing a SET REPLACE=, it was discovered that the container specified already had a replacement policy specified.
Error 3260: is a TRANSPORTABLE disk. TRANSPORTABLE disks cannot be used by storagesets. Do a SET NOTRANSPORTABLE before using this disk in a storageset Explanation: You cannot place a TRANSPORTABLE disk into a reduced RAIDset. Set the disk NOTRANSPORTABLE and retry the command. Error 3270: not in NORMAL state. Only NORMAL state units may be specified as a read source Explanation: You may not specify a MIRRORset member as a read source unless it’s in NORMAL state.
Error 3340: Can only UNMIRROR disks. is not a disk Explanation: Only disks may be unMIRRORed. If you specify a device other than a disk drive, the above error is printed. Try the command again specifying a disk drive. Error 3350 must be configured under a mirrorset to UNMIRROR Explanation: Only disk drives that are configured under mirrorsets may be unMIRRORed. Specify a disk that is configured under a MIRRORset.
Error 3420: is not part of the same unit as previous disks specified Explanation: All disks specified on the REDUCE command must be used by one common unit. If the disks specified are in use by more than one unit, the above message is printed. Retry the command specifying disks that all have one common unit as a parent. Error 3430: A REDUCE may not be done on disks not configured as a unit Explanation: A REDUCE command is only valid when the storageset is configured as a unit.
Error 4010: Illegal character in CLI prompt. Explanation: A nonprintable character was specified. Only ASCII characters space ‘‘ ’’ through tilde ‘‘~’’ may be specified (hex 20–7E). Error 4020: Terminal speed must be 300, 1200, 2400, 4800, 9600 or 19200 Explanation: This error results from a SET THIS_CONTROLLER or SET OTHER_CONTROLLER command with the argument TERMINAL_SPEED=. The only valid baud rates that may be specified are 300, 1200, 2400, 4800, 9600 or 19200 baud.
Error 4090: Module has invalid serial number. This controller cannot be used Call field service Explanation: This error is typically the result of faulty Non-Volatile memory. This error cannot be fixed in the field. A replacement controller must be orderd. Contact Digital Multivendor Customer Services. Error 4100: Unable to RESTART other controller. Explanation: A communication error occurred when trying to restart the other controller. Retry the RESTART command.
Error 4160: Unable to rundown the following units on this controller: Explanation: When attempting to SHUTDOWN, RESTART or SELFTEST this controller, some units could not be successfully run down. This can be caused either by online units or errors when trying to rundown the units. Either rectify the problems on the problem units or issue the SHUTDOWN, RESTART or SELFTEST command with the qualifier OVERRIDE_ONLINE or IGNORE_ERRORS.
Error 5020: The requested program is unknown. Explanation: This error results from a ‘‘RUN .’’ Enter ‘‘DIR’’ to get a list of available programs. Error 5030: Insufficient memory for request. Explanation: This error results from a ‘‘RUN ’’ resource problem. Retry the command later. Error 6000: Communication failure with the other controller. Explanation: There was a communication problem with the other controller. This typically happens if the other controller is shutting down.
Error 6070: Illegal command—this controller not configured for dual-redundancy Explanation: A command was entered to a single controller configuration that requires two controllers to be in dual-redundant mode. If two controllers are supposed to be in dual-redundant mode, enter a SET FAILOVER command. If not, do not enter the command that resulted in the error.
Error 6140: Writeback cache in use on this controller but not enabled on the other controller. Explanation: When trying to SET FAILOVER, it was discovered that there were write-back cache switches set on this controller but the other controller did not have the write-back cache feature enabled. If write-back cache is licensed on the other controller, enable it.
Error 6190: MIRRORING in use on the other controller but not enabled on this controller Explanation: When trying to SET FAILOVER, it was discovered that there were MIRRORsets on the other controller but this controller did not have the MIRROR feature enabled. If MIRROR is licensed on this controller, enable it. If it is not licensed, contact Digital Multivendor Customer Services for licensing information or do not use the two controllers in dual-redundant mode, or do not use a MIRRORset configuration.
Error 7080: Cannot retry UNWRITEABLE_DATA on a unit without UNWRITEABLE_DATA Explanation: If a unit does not have a UNWRITEABLE_DATA error, a RETRY UNWRITEABLE_DATA is an illegal command. Check to assure the unit that you wished to RETRY UNWRITEABLE_DATA on does exhibit a UNWRITEABLE_DATA error. Error 9000: Cannot rename a unit Explanation: Only devices and storagesets may be renamed. If you attempt to rename a unit, the above message results.
Error 9080: is already used Explanation: An ADD command specified a name that is already in use. Specify another name. Note This error is commonly the result of failing to name a storageset.
Error 9170: at PTL No device installed Explanation: When a unit is added or initialized, the configuration of the devices that makes up the unit is checked. If no device is found at the PTL specified, this error is displayed. Check both the logical and physical configuration of the unit and correct any mismatches.
Error 9240: Cannot delete unit in maintenance mode Explanation: When trying to delete a unit, the unit was found to be in maintenance mode. This is typically the result of trying to delete a unit that is in use by DILX or TILX. Ensure that DILX and TILX is not being run against the unit that is to be deleted, and retry the command. Error 9250: Initialize of disk failed Explanation: Unable to write metadata on disk. Make sure the disk is not broken.
Error 9330: NV memory write collision. Please try again Explanation: Two processes were trying to modify the controller’s configuration at the same time. Check the configuration you were trying to modify to make sure it’s unchanged and retry the command. Error 9340: Reduced raidsets cannot be INITIALIZED Explanation: You cannot INITIALIZE a RAIDset that is running in reduced state. Replace a member and try again.
Error 9420: Unit has unflushed data or a cache error and must be deleted on this controller Explanation: When trying to set failover a unit with unflushed data or a cache error was detected on this controller. Delete the unit as requested and then retry the SET FAILOVER command. Error 9430: Cannot check if drives have unflushed data or cache errors on the other controller Explanation: Communication error when trying to SET FAILOVER. Retry the command.
Warning 3000: This storageset is configured with more than one disk per port. This causes a degradation in performance Explanation: This warning results from an ADD storageset-type command. The storageset specified has more than one member per port. One method of increasing the controller’s performance is through parallel transfers to members of a storageset. If multiple members of a storageset are on one port, transfers must be done in serial to those members.
Warning 4020: A restart of both this and the other controller is required before all the parameters modified will take effect Explanation: This warning results from a SET THIS_CONTROLLER or a SET OTHER_CONTROLLER command. Some controller parameters require a restart of both controllers before they can take effect. If any of those parameters are changed, this warning is displayed. Restart both controllers and retry the command.
Warning 7020: Unable to clear UNWRITEABLE_DATA on other controller Explanation: When trying to clear UNWRITEABLE_DATA on the other controller, a communication error occurred. Retry the command. If the failure persists, contact Digital Multivendor Customer Services. Warning 9000: Drive has LOST_DATA Explanation: During a check of the drive’s metadata, it was detected that the drive had lost data. Clear the lost data error on the drive.
Warning 9090: Metadata found on container. Are you sure this is a TRANSPORTABLE container? Explanation: When a transportable disk was initialized, metadata was found. Verify that this disk in fact should be marked transportable. No action is required to correct this warning. Warning 9100: Bad or low battery or bad write cache on writeback cache will not be used Explanation: The battery is low or bad on the specified controller.
Example B–3 Initial Single Controller Configuration of an HSZ Controller CLI> SET THIS_CONTROLLER ID=5 CLI> RESTART THIS_CONTROLLER [this controller restarts at this point] Example B–4 Setting the Terminal Speed and Parity CLI> SET THIS_CONTROLLER TERMINAL_SPEED=19200 NOTERMINAL_PARITY Note Garbage will appear on the terminal after setting the controller’s terminal speed until you set the terminal’s speed to match the controller’s new terminal speed.
Example B–8 Creating a Unit from a Five-Member RAIDset CLI> ADD DISK DISK0 1 0 0 CLI> ADD DISK DISK1 2 0 0 CLI> ADD DISK DISK2 3 0 0 CLI> ADD DISK DISK3 1 1 0 CLI> ADD DISK DISK4 2 1 0 CLI> ADD RAIDSET RAID9 DISK0 DISK1 DISK2 DISK3 DISK4 Warning 3000: This storageset is configured with more than one disk per port.
Example B–13 Changing the Replacement Policy of a RAIDset CLI> CLI> CLI> CLI> CLI> CLI> CLI> CLI> CLI> ADD DISK DISK0 1 0 0 ADD DISK DISK1 2 0 0 ADD DISK DISK2 3 0 0 ADD DISK DISK3 4 0 0 ADD DISK DISK4 5 0 0 ADD RAIDSET RAID9 DISK0 DISK1 DISK2 DISK3 DISK4 INITIALIZE RAID9 ADD UNIT D0 RAID9 SET RAID9 POLICY=BEST_FIT Example B–14 Deleting the Unit, Stripeset, and All Disks Associated with a Stripeset CLI> CLI> CLI> CLI> CLI> CLI> DELETE DELETE DELETE DELETE DELETE DELETE D0 STRIPE0 DISK0 DISK1 DISK2 DISK3
Glossary ac distribution The method of controlling ac power in a cabinet. adapter A device that converts the protocol and hardware interface of one bus type into that of another without changing the functionality of the bus. See SCSI bus signal converter. allocation class A numerical value assigned to an HSJ or HSD30 controller to uniquely identify units across multiple, independent controllers. (Controllers in a dual-redundant configuration must have the same allocation class.
CDU Cable distribution unit. The power entry device for StorageWorks cabinets. The unit provides the connections necessary to distribute ac power to cabinet shelves and fans. CI bus Computer interconnect bus. Uses two serial paths, each with a transfer rate of 70 MB/s (8.75 MB/s). CLI Command line interpreter. Operator command line interface for the HS family controller firmware. container Any entity that is capable of storing data, whether it is a physical device or a group of physical devices.
dual-redundant Two controllers in one controller shelf providing the ability for one controller to take over the work of the other controller in the event of a failure of the other controller. DUART Dual universal asynchronous receiver transmitter. An integrated circuit containing two serial, asynchronous transceiver circuits. DUP Diagnostic and Utility Protocol. Host application software that allows a host operator terminal to connect to the controller’s command line interpreter. See also virtual terminal.
half-height device A device that occupies half of a 5.25 inch SBB carrier. Two half-height devices can be mounted in a 5.25 inch SBB carrier. The first half-height device is normally mounted in the lower part of the carrier. The second device is normally mounted in the upper part of the carrier. HBVS Host-Based Volume Shadowing. Also known as Phase 2 Volume Shadowing. HSOF HSOF. An abbreviation for Hierarchical Storage Operating Firmware. HIS Host Interconnect Services.
LUN A logical unit number is a physical or virtual peripheral device addressable through a target. LUNs use their target’s bus connection to communicate on the SCSI bus. maintenance terminal Any EIA–423 compatible terminal to be plugged into the HS controller. This terminal is used to identify the controller, enable host paths, define the configuration, and check controller status. It is not required for normal operations. It is sometimes referred to as a local terminal.
port The hardware and software used to connect a host controller to a communication bus, such as CI, DSSI, or SCSI bus. This term also is used to describe the connect between the controller and SCSI storage devices. PTL Port-Target-LUN device notation. Where P designates the port (1 through 6), T designates the target ID of the device (0 through 6 in a nonfailover configuration, or 0 through 5 if the controller is in a failover configuration), and L designates the LUN of the device (0 through 7).
SCSI Small Computer System Interface. An ANSI interface defining the physical and electrical parameters of a parallel I/O bus used to connect hosts to a maximum of seven devices. The StorageWorks device interface is implemented according to SCSI–2 standard, allowing the synchronous transfer of 8-bit data at rates of up to 10 MB/s. SCSI device A host computer adapter, a peripheral controller, or a storage element that can be attached to the SCSI bus.
striped mirrorsets Stripesets whose members have been mirrored. tagged command queuing A technique that allows a device to have multiple I/O requests outstanding to it at one time. target Is a SCSI device that performs an operation requested by an initiator. Target is determined by the device’s address on its SCSI bus. For example, the HSJ controller can address targets 0 through 6 in a single configuration or targets 0 through 5 in a dual-redundant configuration. TILX Tape Inline Exerciser.
write-back A cache write strategy that writes to the cache memory, the MAY flush the data to the primary media at some future time. The user sees the operation as complete when the data has reached the cache. The intent of this strategy is to avoid unnecessary accesses to the primary media. write hole Undetectable RAID level 1 or 5 data corruption. A write hole is caused by the successful writing of some, but not all, of the storageset members.
Index A Acceptance tests with power applied, 5–17 Adapter support for HSD30 array controllers, 3–25 support for HSJ array controllers, 3–25 support for HSZ40 array controllers, 3–26 ADD CDROM command, B–2 ADD DISK command, B–3 ADD LOADER command, 7–101, B–5 ADD MIRRORSET command, B–6 ADD OPTICAL command, B–8 ADD PASSTHROUGH command, 7–101, B–10 ADD RAIDSET command, B–11 ADD SPARESET command, B–13 ADD STRIPESET command, B–14 ADD TAPE command, B–15 ADD UNIT command, B–16 Adding a device warm swap, 5–29 Adding
CACHE_POLICY=B qualifier for RAIDsets for mirrorsets, 2–4 CD–ROM always set TRANSPORTABLE, 5–9 configuration restrictions, 3–25 CFMENU utility, 7–59 devices, 7–60 example, 7–59 exiting, 7–71 failedset, 7–64 fields, 7–61, 7–62, 7–63, 7–64, 7–65, 7–67 initializing containers, 7–66 invoking, 7–60 main menu, 7–60 messages, 7–68 mirrorsets, 7–62 passthrough containers, 7–65 RAIDset, 7–63 restrictions, 7–60 spareset, 7–64 storageset, 7–62, 7–63, 7–64 stripeset, 7–62 terminal setup, 7–68 units, 7–67 Chunksize defa
Commands (cont’d) ADD UNIT, B–16 CLEAR_ERRORS CLI, B–27 CLEAR_ERRORS INVALID_CACHE, 6–9, B–28 CLEAR_ERRORS LOST_DATA, B–29 CLEAR_ERRORS UNKNOWN, B–30 CLEAR_ERRORS UNWRITEABLE_DATA, 6–8, B–31 DELETE container-name, B–32 DELETE FAILEDSET, B–33 DELETE SPARESET, B–34 DELETE unit-number, B–35 DIRECTORY, B–36 disk-device-name, B–129 EXIT, B–37 HELP, B–38 INITIALIZE, B–39 LOCATE, B–41 LOCATE CANCEL, B–41 LOCATE DISKS, B–41 LOCATE entity, B–42 LOCATE LOADERS, B–41 LOCATE OPTICALS, B–41 LOCATE PTL SCSI-location, B–4
Configuration menu See CFMENU utility Configuration mismatch port LED state, 5–23 Configuration parameters Are they preset?, 4–20 Configuration rules for dual-redundant, 3–13 for SCSI–2 device cables, 4–9 for shelf power, 4–2 for SW300-series cabinet, 3–8 for SW300-series cabinet shelf, 3–8 for SW500-series cabinet, 3–6 for SW800-series cabinet, 3–2 for SWxxx cabinets, 4–13 nonredundant controller, 3–13 shelf, 3–10 SW300-series cabinet shelf, 3–11 SW800-series SW500-series, 4–13 Configure-to-order See CTO C
DILX (cont’d) invoking for virtual terminal DUP connection, 7–3 invoking from maintenance terminal, 7–3 invoking from VCS, 7–3 User-Defined test, 7–6 with command disks, 7–101 DIRECTORY command, B–36 Disk drive removal from mirrorsets, 5–28 warm swap, 5–28, 5–29 warm swap removal, 5–28 warm swap replacement, 5–29 DISMOUNT command (host) use of, 5–45 DSSI bus rules, 1–6 DSSI bus termination, 1–6 DSSI host port cables, 3–24, 4–7 lengths connector type, 4–8 node-to-node max. length end-to-end max.
EXIT command, B–37 from CLI, 5–6 F Failback operation from failover, 2–22 Failedset CFMENU utility, 7–64 delete, 6–14 showing members deleting member from, 6–14 Failover, 2–12 as seen by HSZ hosts, 2–22 configuration, 2–16 correcting mismatch, 2–19 description of, 2–16 exiting, 2–18 failback operation, 2–22 for HSZ40, 2–20 HSZ host bus, 2–18 initiate with C_SWAP, 2–22 KILL signal, 2–20 RAIDsets, 2–4 resolving a hardware mismatch, 2–19 restart mismatch, 2–19 reviving failed controller, 2–19 setting of, 2–17
HS array controller (cont’d) executive functions, 2–13 features of, 1–12 firmware overview, 2–12 hardware functional overview, 2–1 initialized by, 5–1 interface host port, 2–3 OpenVMS AUTOGEN.
L Lamp test, B–41 LED error codes, 5–20 how to clear, 5–24 on OCP, 2–6 LEDs OCP, 5–18 License keys FLS, 7–54 Licensed features, 7–54 hourly message, 6–9 Loaders, 7–101 Local programs, 2–14, 7–2 LOCATE CANCEL command, B–41 LOCATE command, B–41 LOCATE DISKS command, B–41 LOCATE entity, B–42 LOCATE LOADERS command, B–41 LOCATE OPTICALS command, B–41 LOCATE PTL SCSI-location command, B–42 LOCATE TAPES command, B–41 LOCATE UNITS command, B–42 Log file from virtual terminal, 5–6 /LOG= qualifier, 8–3 Logical unit
O P OCP description of, 2–6, 5–18 disk drive configuration mismatch disk drive failure, 5–23 for HSD30 array controller, 2–7 for HSJ array controllers, 2–7 for HSZ array controllers, 2–8 how it functions, 5–18 LED code normal state, 5–24 LED error codes, 5–20 Led error codes, clearing, 5–24 use of buttons, 5–18 use of LEDs, 5–18 OpenVMS preferred path utility, 8–6 SHOW DEVICE, 8–6 OpenVMS Alpha DECevent, 8–13 ERF support, 8–9 loading the FYDRIVER, 8–16 OpenVMS VAX AUTOGEN.
Protocols MSCP (cont’d) TMSCP, 1–19 SCSI (to devices), 1–19 SCSI (to hosts), 1–19 PTL controller perspective, 1–10 host perspective, HSZ40 array controllers, 1–12 specifying device locations, 5–5 Q Quiesce the bus, 5–27 troubleshooting, 5–30 Quiet slot time setting for HSJ, 4–15 R RAID FLS, 7–54 levels supported, 1–13 RAID 0 striping, 6–2 RAID 0 & 1, 6–2 RAID 1 mirroring, 6–2 RAID 3 elements, 6–3 RAID 5, 6–4 Turning on the license key, 7–54 RAID 0 ADD STRIPESET, 6–17 SHOW STRIPESET, 6–17 RAIDset ADD RAIDS
REPLACE=disk-device-name, B–62 Replacement read cache write-back cache, 5–43 Replacing a device warm swap, 5–29 Reset (//) button, 5–18 controller failure, 2–6 controller initialization, 2–19 controller restarts, 2–6 Reset state recovering from, 2–10 RESTART OTHER_CONTROLLER command, 2–19, B–48 RESTART THIS_CONTROLLER command, B–50 Restrictions CFMENU utility, 7–60 FLS, 7–54 RETRY_ERRORS UNWRITEABLE_DATA command, B–53 Rules for system power off, 5–43 RUN command, B–54 S SBB shelf, 1–18 SBB shelf device 3½-
SHOW stripeset-container-name command, B–117 SHOW STRIPESETS command, B–115 SHOW tape-container-name command, B–119 SHOW TAPES command, B–118 SHOW THIS_CONTROLLER command, B–120 determining the cache module type, 4–31 SHOW unit-number command, B–124 SHOW UNITS command, B–123 SHUTDOWN controller, 5–43 use of, 5–45 SHUTDOWN OTHER_CONTROLLER command, 5–43, B–125 SHUTDOWN THIS_CONTROLLER command, 5–43, B–127 Site preparation, 4–1 Slot numbers for controllers, 1–18 for SBBs, 1–18 Sparesets adding members deletin
TMSCP protocol for HSJ and HSD30, 1–19 Tools for controller module removal, 5–35 for subsystem installation, 4–3 Transportable, 5–17 adding non-HS array controller devices, 5–17 TRANSPORTABLE qualifier CLI, 5–10 Trilink connector block for HSD30 array controllers, 1–5 for HSZ40 array controllers, 1–7 installation of (HSZ), 4–23 installation on an HSD30, 4–22 part number for HSD30, 4–22 part number for HSZ, 4–23 U Units adding, B–16 CFMENU utility, 7–67 UNMIRROR disk-device-name command, B–129 Unpacking sub
