HP Integrity Virtual Machines 4.2: Installation, Configuration, and Administration

Dynamic memory is an optional feature of Integrity VM that allows you to change the amount
of physical memory in use by a virtual machine without rebooting the virtual machine. In VMs
as Serviceguard Package configurations, the virtual machines can be configured with differing
dynamic memory configurations on different Serviceguard nodes. If you plan to use the dynamic
memory feature, you should understand how it might impact your virtual machine configurations
and failover strategy. For details about configuring dynamic memory, see Section 9.7 (page 151).
11.3.3 Network Components
To maximize availability, a minimum of three physical network interface cards (pNIC) are
recommended, one each VM Host. One pNIC is configured as a dedicated Serviceguard heartbeat,
and therefore, is not available for any use by Integrity Virtual Machines. Two additional pNICs
are configured as a primary or standby LAN pair for use by VM guests and are monitored by
Serviceguard on the VM Host. The Serviceguard network monitor provides network failure
detection for identifying failed network cards based on inbound and outbound message counts
and failing over to configured standby LANs.
Auto-Port Aggregation (APA) is supported and can be used to provide network bandwidth
scalability, load balancing between the physical links, automatic fault detection and HA recovery.
NOTE: When using APA, make sure you have at least two pNICs configured to avoid a
single-point-of-failure for the Serviceguard cluster heartbeat connections.
As with all VM configurations, to allow guests to access network devices, you must create
vswitches on the VM Host. When configuring vswitches in a VMs as Serviceguard Packages
configuration, you configure vswitches only on the pNICs associated with the primary LANs.
You do not configure a vswitches on the pNIC associated with the Serviceguard Heartbeat
network or the pNIC associated with the standby LAN.
When the Serviceguard network monitor detects a network failure on the pNIC associated with
a primary LAN, it fails the LAN over to the pNIC associated with the standby LAN. When this
occurs, the vswitch monitor, which is responsible for monitoring the activities of the Serviceguard
network monitor, automatically moves the vswitch configuration from the pNIC associated with
the primary LAN to the pNIC associated with the standby LAN.
All VLAN configurations supported by Integrity Virtual Machines are supported in VMs as
Serviceguard Packages configurations. All VLAN configurations supported by Serviceguard are
supported in VMs as Serviceguard Packages configurations.
11.3.4 Storage Considerations
VMs as Serviceguard Package configurations support all VM guests backing store types, including:
• Whole disks
• LVM logical volumes
• VxVM logical volumes
• Files on any of the storage types listed, including files on a Cluster File System (CFS).
The VM guest backing stores must reside on shared storage, so that it is accessible by all VM
Hosts in the Serviceguard cluster to allow failover of the VM guests. Configuring shared storage
for VM Hosts is accomplished in a similar manner as a standard Serviceguard cluster
configuration, with the only difference being the shared storage must be defined as VM storage
devices that are exclusively used by a specific VM guest.
Both standard LVM and VxVm logical volumes are only active and available on one Serviceguard
node at a time. Whole disks, Shared LVM logical volumes, Vxvm logical volumes managed by
the Cluster Volumes Manager (CVM) as well as files on a Cluster Files System can be configured
to be active and available on multiple Serviceguard nodes simultaneously. Additional care and
planning should be taken when configuring VMs as Serviceguard Packages to insure that no
more then one guest on one node can access these backing store types.
196 Using HP Serviceguard with Integrity VM