Carrier Grade Dell Chassis Management Controller Firmware User’s Guide Addendum
Notes and Cautions NOTE: A NOTE indicates important information that helps you make better use of your computer. CAUTION: A CAUTION indicates potential damage to hardware or loss of data if instructions are not followed. ____________________ Information in this publication is subject to change without notice. © 2012 Dell Inc. All rights reserved. Reproduction of these materials in any manner whatsoever without the written permission of Dell Inc. is strictly forbidden.
Contents Overview . . . . . . . . . . . . . . . . . . . . . . . . . . Power Management . . . . . . . . . . . . . . . . . . . . 5 5 Grid Redundancy Mode . . . . . . . . . . . . . . . 6 Grid Redundancy Levels . . . . . . . . . . . . . . . 6 Power Supply Redundancy Mode No Redundancy Mode . . . . . . . . . . 7 . . . . . . . . . . . . . . . . 8 Power Budgeting for Hardware Modules Server Slot Power Priority Settings . . . . . . . . . . . . . . . . . . . . . . .
Power Supply and Redundancy Policy Changes in System Event Log . . . . . . . . . . . . . . . . . . . 23 . . . . . 25 . . . . . . . . . . . 25 Redundancy Status and Overall Power Health Configuring and Managing Power. . . . . . . 25 . . . . . . . . 28 . . . . . . . . . . . 33 Viewing the Health Status of the PSUs . Viewing Power Consumption Status Viewing Power Budget Status Configuring Power Budget and Redundancy . . . . . . . . . . . . . . . . . . . . .
Overview This document provides additional information for the Carrier Grade Dell Chassis Management Controller when running with DC input power supply units in a Network Equipment-Building Standards (NEBS) configuration. The information presented in this addendum supersedes the information as presented in the Dell CMC Controller Firmware Version 4.1 User's Guide. For more information, see the CMC Online Help for Carrier Grade CMC.
For more information, see the External Power Management section in the Chassis Management Controller Version 4.1 User’s Guide. NOTE: Actual power delivery is based on configuration and workload. The Power Management features of the M1000e help administrators configure the enclosure to reduce power consumption and to customize power management to their unique requirements and environments.
for some reason CMC is unable to maintain grid redundancy, then E-mail and/or SNMP alerts are sent to administrators if the Redundancy Lost event is configured for alerting. Figure 1-1.
NOTE: Dynamic Power Supply Engagement (DPSE) allows PSUs to be placed in standby. The standby state indicates a physical state: that of not supplying power. When you enable DPSE, the extra PSUs may be placed in Standby mode to increase efficiency and save power. Figure 1-2. Power Supply Redundancy: Totally 4 PSUs with a failure of one PSU.
Figure 1-3. No Redundancy with three PSUs in the chassis DC Power Grid #1 Power Supply #1 Power Supply #2 Power Supply #3 Empty Slot #4 Empty Slot #5 Empty Slot #6 Chassis DC Power Bus Single Power Grid: No protection against grid or power supply failure A PSU failure brings other PSUs out of Standby mode, as needed, to support the chassis power allocations. If you have 4 PSUs, and require only three, then in the event that one fails, the fourth PSU is brought online.
Figure 1-4. Chassis With Six-PSU Configuration PSU1 PSU2 PSU3 PSU4 PSU5 PSU6 CMC maintains a power budget for the enclosure that reserves the necessary wattage for all installed servers and components. CMC allocates power to the CMC infrastructure and the servers in the chassis. CMC infrastructure consists of components in the chassis, such as fans, I/O modules, and iKVM (if present). The chassis may have up to 32 servers that communicate to the chassis through the iDRAC.
CMC grants the requested power to the server, and the allocated wattage is subtracted from the available budget. Once the server is granted a power request, the server's iDRAC software continuously monitors the actual power consumption. Depending on the actual power requirements, the iDRAC power envelope may change over time. iDRAC requests a power step-up only if the servers are fully consuming the allocated power.
Additional servers can be powered up in the modular enclosure only if sufficient power is available. The System Input Power Cap can be increased any time up to a maximum value of 16685 watts to allow the power up of additional servers. Changes in the modular enclosure that reduce the power allocation are: • Server power off • Server • I/O module • iKVM removal • Transition of the chassis to a powered off state You can reconfigure the System Input Power Cap when chassis is either ON or OFF.
to accommodate the higher priority servers. So after the available power for allocation is exhausted, then CMC reclaims power from lower or equal priority servers until they are at their minimum power level. NOTE: I/O modules, fans, and iKVM (if present) are given the highest priority. CMC reclaims power only from lower priority devices to meet the power needs of a higher priority module or server. Dynamic Power Supply Engagement Dynamic Power Supply Engagement (DPSE) mode is disabled by default.
• In a No Redundancy configuration with DPSE, the M1000e can have up to five power supply units in Standby state. In a six PSU configuration, some PSU units are placed in Standby and are not utilized to improve power efficiency. Removal or failure of an online PSU in this configuration cause a PSU in Standby state to change to Online; however, standby PSUs can take up to two seconds to become active, so some server modules may lose power during the transition in the No Redundancy configuration.
Redundancy Policies Redundancy policy is a configurable set of properties that determine how CMC manages power to the chassis. The following redundancy policies are configurable with or without dynamic PSU engagement: • Grid Redundancy • Power Supply Redundancy • No Redundancy The default redundancy configuration for a chassis depends on how many PSUs it contains, as shown in Table 1-1. Table 1-1.
No Redundancy Power in excess of what is necessary to power the chassis is available, even on a failure, to continue to power the chassis. CAUTION: The No Redundancy mode uses optimum PSUs when DPSE is enabled for the requirements of the chassis. Failure of a single PSU could cause servers to lose power and data in this mode. Power Conservation and Power Budget Changes CMC performs power conservation when the user-configured maximum power limit is reached.
In maximum power conservation mode, all servers start functioning at their minimum power levels, and all subsequent server power allocation requests are denied. In this mode, the performance of powered on servers may be degraded. Additional servers cannot be powered on, regardless of server priority. The system is restored to full performance when the user or an automated command line script clears the maximum conservation mode.
• Granted sufficient power for full performance. • Powered on. Using Web Interface To enable Server Performance Over Power Redundancy, perform the following steps: 1 Click Chassis Overview in the system tree. 2 Click Power Configuration. 3 Select Server Performance Over Power Redundancy and click Apply. Using RACADM To enable Server Performance Over Power Redundancy, perform the following steps: 1 Open a serial, Telnet, or SSH text console to CMC and log in.
Using Web Interface You can enable power remote logging using the GUI. To do this, log in to the GUI, and do the following: 1 Click Chassis Overview in the system tree. 2 Click Power Configuration. 3 Select Power Remote Logging, to enable you to log power events to a remote host. 4 Specify the required logging interval (1-1440 minutes). 5 Click Apply to save changes.
NOTE: The power remote logging feature is dependent on remote syslog hosts having been previously configured. Logging to one or more remote syslog hosts must be enabled, otherwise power consumption is logged. This can be done either through the Web GUI or the RACADM CLI. For more information, see the remote syslog configuration instructions. PSU Failure With Degraded or No Redundancy Policy CMC decreases power to servers when an insufficient power event occurs, such as a PSU failure.
Table 1-2 describes the actions taken by CMC when a new server is powered on in the scenario described earlier. Table 1-2.
Table 1-3. Chassis Impact from PSU Failure or Removal PSU Configuration Dynamic PSU Engagement Firmware Response Grid Redundancy Enabled CMC alerts you of loss of Grid Redundancy. PSUs in standby mode (if any) are turned on to compensate for power budget lost from the PSU failure or removal. Power Supply Redundancy Enabled CMC alerts you of loss of Power Supply Redundancy. PSUs in standby mode (if any) are turned on to compensate for power budget lost from PSU failure or removal.
• Changes to the PSU redundancy policy are limited by the number of PSUs in the chassis. You can select any of the three PSU redundancy configuration settings. For more information, see Redundancy Policies. Power Supply and Redundancy Policy Changes in System Event Log Changes in the power supply state and power redundancy status are recorded as events.
Table 1-4. SEL Events for Power Supply Changes Power Supply Event System Event Log (SEL) Entry Input under-voltage An under voltage fault detected on power supply . Input over-current An over current fault detected on power supply . Input under-current An under current fault detected on power supply . DC output over-voltage An output over voltage fault detected on power supply .
Table 1-4. SEL Events for Power Supply Changes Power Supply Event System Event Log (SEL) Entry Other fault Power supply failed. Redundancy Status and Overall Power Health The redundancy status is a factor in determining the overall power health. When the power redundancy policy is set, for example, to Grid Redundancy and the redundancy status indicates that the system is operating with redundancy, the overall power health is typically OK.
The Chassis Graphics page provides a graphical overview of all PSUs installed in the chassis. To view health status for all PSUs using Chassis Graphics: 1 Log in to the CMC Web interface. The Chassis Status page is displayed. The lower section of Chassis Graphics depicts the rear view of the chassis and contains the health status of all PSUs.
Table 1-5. Power Supplies Item Description Name Displays the name of the power supply unit: PS-[n], where [n] is the power supply number. Present Indicates whether the PSU is Present or Absent. Health OK Indicates that the PSU is present and communicating with CMC. In the event of a communication failure between CMC and the power supply, CMC cannot obtain or display health status for the PSU. Warning Indicates that only Warning alerts have been issued, and corrective action must be taken.
Table 1-6. System Power Status (continued) PSU Configuration Dynamic PSU Engagement Redundancy Displays the power supply redundancy status. Values include: No: Power Supplies are not redundant. Yes: Full Redundancy in effect. Using RACADM Open a serial/Telnet/SSH text console to CMC, log in, and type: racadm getpminfo For more information about getpminfo, including output details, see the RACADM Command Line Reference Guide for iDRAC7 1.00.00 and CMC 4.
Table 1-7. System Power Status Item Description Overall Power Health Indicates the health status of the chassis' power subsystem: • Green check icon for OK • Yellow exclamation icon for non-critical • Red x icon for critical System Power Status Displays the power status (On, Off, Powering On, Powering Off) of the chassis. Redundancy Displays the redundancy status. Valid values are: • No - PSUs are not redundant. • Yes - Full redundancy in effect. Table 1-8.
Table 1-8. Real-Time Power Statistics (continued) 30 Item Description Peak System Power Timestamp Displays the minimum system level input power consumption value (in watts) over the time since the user last cleared this value. This property allows you to track the minimum power consumption by the system (chassis and modules) recorded over a period of time. Click Reset Peak/Min Power Statistics below the table to clear this value.
Table 1-8. Real-Time Power Statistics (continued) Item Description System Idle Power Displays the estimated power consumption of the chassis when it is in idle state. The idle state is defined as the state of the chassis while it is ON and all modules are consuming power while in the idle state. This is an estimated value and not a measured value.
Table 1-9. Real-Time Energy Statistics Status Item Description System Energy Consumption Displays the current cumulative energy consumption for all modules in the chassis measured from the input side of the power supplies. The value is displayed in KWh and it is a cumulative value. System Energy Consumption Start Time Displays the date and time recorded when the system energy consumption value was last cleared, and the new measurement cycle began.
Table 1-10. Server Modules Item Description Cumulative Power Start Time Real-time measurement of the cumulative power that the server has consumed since the time displayed in the Start Time field. The measurement is presented in kilowatt hours (kWh). Peak Consumption Time Stamp Displays the peak power that the server consumed at one time. The time when the peak power consumption occurred is recorded in the Time Stamp field. The measurement is displayed in watts.
Table 1-11. 34 System Power Policy Configuration Item Description System Input Power Cap Displays the user configured maximum power consumption limit for the entire system (chassis, CMC, servers, I/O modules, power supply units, iKVM, and fans). CMC enforces this limit via reduced server power allocations, or by powering off lower priority server modules. The value for system input power cap is displayed in watts, BTU/h and percent units.
Table 1-11. System Power Policy Configuration Item Description Redundancy Policy Displays the current redundancy configuration: AC Redundancy, Power Supply Redundancy, and No Redundancy. • Grid Redundancy - Power input is load-balanced across all PSUs. Half of them should be cabled to one power grid and the other half should be cabled to another grid. When the system is running optimally in Grid Redundancy mode, power is loadbalanced across all active supplies.
Table 1-12. Power Budgeting Item Description System Input Max Power Capacity Maximum input power that the available power supplies can supply to the system (in watts). Input Redundancy Reserve Displays the amount of redundant power (in watts) in reserve that can be utilized in the event of an input power grid or power supply unit (PSU) failure.
Table 1-13. Server Modules Item Description Slot Displays the location of the server module. The Slot is a sequential number (1-16) that identifies the server module by its location within the chassis. Name Displays the server name. The server name is defined by the user. Type Displays the type of the server. Priority of the relocated server. Displays the priority level allotted to the server slot in the chassis for power budgeting.
Table 1-14. Chassis Power Supplies Item Description Name Displays the name of the PSU in the format PS-n, where n, is the PSU number. Power State Displays the power state of the PSU - Initializing, Online, Stand By, In Diagnostics, Failed, Unknown, or Absent (missing). Input Volts Displays the present input voltage of the power supply. Input Current Displays the present input current of the power supply. Output Rated Power Displays the maximum output power rating of the power supply.
Table 1-15. Item Configurable Power Budget/Redundancy Properties Description System Input Power Cap System Input Power Cap is the maximum input power that the system is allowed to allocate to servers and chassis infrastructure. It can be configured by the user to any value that exceeds the minimum power needed for servers that are powered on and the chassis infrastructure; configuring a value that falls below the minimum power needed for servers and the chassis infrastructure fails.
Table 1-15. Configurable Power Budget/Redundancy Properties Item Description Redundancy Policy This option allows you to select one the following options: • No Redundancy: Power from the power supplies is used to power the entire chassis, including the chassis, servers, I/O modules, iKVM, and CMC. No power supplies must be kept in reserve. NOTE: The No Redundancy mode uses only the minimum required number of power supplies at a time.
Table 1-15. Configurable Power Budget/Redundancy Properties Item Description Server Performance Over This option favors server performance and server power up, Power Redundancy over maintaining power redundancy. For more information about this feature, see Server Performance Over Power Redundancy. Enable Dynamic Power Supply Engagement On selection, enables dynamic power management.
Using RACADM To enable and set the redundancy policy: NOTE: To perform power management actions, you must have Chassis Configuration Administrator privilege. 1 Open a serial/Telnet/SSH text console to CMC and log in. 2 Set properties as needed: • To select a redundancy policy, type: racadm config -g cfgChassisPower -o cfgChassisRedundancyPolicy where is 0 (No Redundancy), 1 (Grid Redundancy), 2 (Power Supply Redundancy). The default is 0.
Assigning Priority Levels to Servers Server priority levels determine which servers the CMC draws power from when additional power is required. NOTE: The priority you assign to a server is linked to its slot and not to the server itself. If you move the server to a new slot, you must reconfigure the priority for the new slot location. NOTE: To perform power management actions, you must have Chassis Configuration Administrator privilege.
Setting Power Budget NOTE: To perform power management actions, you must have Chassis Configuration Administrator privilege. Using the Web Interface To set the power budget using the CMC Web interface: 1 Log in to the CMC Web interface. 2 Click Chassis Overview in the system tree. The Chassis Health page is displayed. 3 Click the Power tab. The Power Monitoring Status page appears. 4 Click the Configuration subtab. The Budget/Redundancy Configuration page is displayed.
racadm config -g cfgChassisPower -o cfgChassisPowerCap 5400 sets the maximum power budget to 5400 watts. NOTE: The power budget is limited to 16685 Watts. If you attempt to set an input power budget value that exceeds the power capacity of your chassis, CMC displays a failure message.
Carrier Grade Chassis Management Controller User’s Guide Addendum