HP Insight Recovery 1.0 Release Notes

Configuring HP SMI-S EVA to Discover HP Command View EVA Arrays section in the Command
View EVA Installation Guide for details on the Discoverer command.
• HP Insight Recovery 1.0 does not support Command View server credentials that are domain
controller based.
• HP Insight Recovery 1.0 with dissimilar EVA array models between the Primary Site and
Recovery Site, requires the ESX advanced configuration setting Lvm.DisallowSnapshotLun
to be 0 on all ESX servers that are used to host HP Insight Recovery protected Logical servers.
Additionally, HP Virtual Machine Management version 3.7 is required.
• To configure a VM based Logical Server in an HP IR Recovery Group, the VM must be
created using a Logical Server Create operation in Logical Server Management (LSM). VMs
created separately (for example, using VMware Virtual Center), and then imported as Logical
Servers, are not supported. This restriction is related to the issue - Service Restart is
Required if UUID of Virtual Machine Changes During Logical Server
Reactivation, documented in the Insight Dynamics - VSE for ProLiant Version 4.1 Release
Notes.
To resolve this issue, HP recommends upgrading to VMware vCenter Server (formerly
called VMware Virtual Center) version 2.5 Update 4, which contains a fix.
• The configuration of HP Insight Recovery 1.0 at two separate sites is not automatically
synchronized, however HP Insight Recovery provides a configuration export and import
facility that simplifies this task. The HP IR online help will guide you through the export
and import process.
• HP Insight Recovery Version 1.0 supports only StorageWorks EVA storage and Continuous
Access EVA replication in synchronous mode only.
• All of the DR Protected logical servers at a site are failed over; it is not possible to select an
individual logical server, or a subset of DR Protected logical servers, to failover.
• The maximum distance between the Primary and Recovery Site is restricted to metropolitan
area distances. This is related to Fiber Channel SAN design restrictions necessary to support
synchronous replication in Continuous Access EVA.
• To successfully failover a logical server running on a physical (blade) server from the Primary
Site to the Recovery Site, there must be an available blade with a compatible hardware
configuration at the Recovery Site. The requirements of hardware compatibility are the same
for inter-site as for intra-site mobility, and are documented in the HP Virtualization Manager
and logical server Management Version 4.1 Getting Started Guide.
• Excluding the logical server names, the primary and recovery logical servers must be
configured with identical parameters using the ID-VSE logical server manager. There is a
potential for discrepancies, between the Primary and Recovery Sites, in the values of attributes
that are not included in the "logical server Create" screens of the HP IR user interface. You
should assess the impact of these discrepancies on any licensing arrangements in use for
the operating system and applications running on DR Protected logical servers. For example:
— MAC Address - For physical logical servers, the MAC address is assigned from Virtual
Connect or Virtual Connect Enterprise Manager. Disjoint address ranges must be used
at the two sites, hence the MAC address for a primary logical server and the
corresponding recovery logical server will be different.
— HBA WWN - For a physical logical server, the HBA WWN address is assigned via
Virtual Connect or Virtual Connect Enterprise Manager. Disjoint address ranges must
be used at the Primary and Recovery Sites, hence the WWN for a primary logical server
and the corresponding recovery logical server will be different.
— BIOS UUID - There is no supported mechanism to preserve the UUID for a physical
logical server as it moves across sites (as there is when it moves within a site).
Product Issues and Limitations 17