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

the package should be repacked using the hpvmsg_package command, and then re-applied
using the cmapplyconf files as previously outlined.
Maintenance Mode is enabled only on the local node, and when enabled, will periodically note
that it is enabled for each VM in the system syslog file. In the event that the Serviceguard node
on which the VM is running fails during maintenance mode, the VM package will be failed over
to a secondary node. Depending on where in the process the failure occurred, changes made to
the VM or package configuration might or might not have been distributed to the secondary
node prior to the failover. After recovery, ensure that the VM package is not in maintenance
mode and that the configuration is consistent of all nodes.
11.3.11 Serviceguard Virtual Machine Application Monitoring
Starting with Serviceguard A.11.19 and Integrity VM B.04.10 releases, HP supports the monitoring
and control of applications within a virtual machine configured as a Serviceguard package. This
functionality is available for any application running in an HP-UX, Linux, or Windows guest
using the Java Runtime Enviroment (JRE) and provides the following benefits:
• Checks the status of applications within virtual machines from the VM Host under control
of Serviceguard.
• Provides startup and failure detection capabilities for monitored virtual machine applications
using Serviceguard functionality.
• Provides a supported application monitoring framework without having to rely on
custom-written software.
The hpvmsg_package script can help you develop guest applications monitors.
11.4 Combining the VMs as Serviceguard Package and Node Models
VMs as Serviceguard Packages and Serviceguard nodes in separate clusters can co-exist on the
same VM Hosts, allowing failover of the VMs and application packages within their respective
clusters. The combined configuration provides the most flexibility in meeting recovery-time
objectives (RTOs) for VMs and applications while efficiently consolidating systems on VM Hosts.
For example, when consolidating a system environment consisting of Windows servers and
mission-critical applications running on HP-UX servers, the Windows servers with a lower RTO
can be made highly available by running them as VMs encapsulated within Serviceguard packages.
At the same time, the mission-critical applications with a higher RTO can be protected by
Serviceguard packages running within HP-UX VMs, all running together on a minimum number
of VM Hosts.
NOTE: Different versions of Serviceguard can be used for either cluster as long as they are
listed as supported. For a list of supported combinations of Integrity VM and Serviceguard, see
the HP Integrity Virtual Machines 4.2: Release Notes.
When implementing this configuration, there is a possibility that a failure of all heartbeat links
between both clusters could cause the VM node cluster to fail depending on the timing in which
the surviving nodes in each cluster achieves quorum and which cluster reforms first. This failure
scenario can be mitigated by:
• Having multiple redundant heartbeat networks for both cluster
• Using Serviceguard A.11.19 for both clusters and setting the MEMBER_TIMEOUT parameter
on the VMs as Serviceguard Node cluster to at least 1 second greater than the reported
max_reformation_duration value of the VM Host cluster. To determine this value, use
the following command:
cmviewcl -v -f line | grep max_reformation_duration
200 Using HP Serviceguard with Integrity VM