Designing High Availability Solutions with HP Serviceguard and HP Integrity Virtual Machines

13
Figures 6 and 7 are examples showing how Integrity VM can be used to consolidate nodes within Serviceguard
clusters. In figure 7, a single standby host is used to run two separate VM guests that are each part of two different
Serviceguard clusters. The packages that are normally running on primary nodes 1 and 2 can failover to their
corresponding VM guest failover nodes running on a single standby host configured to support the execution of
multiple VM guests.
Figure 7: Serviceguard cluster using VM guests as standby failover nodes
This configuration has the benefit of being able to consolidate standby systems that are normally found in typical
single Serviceguard cluster configurations, thus providing cost savings by reducing the number of required standby
failover systems and making more efficient use of system hardware. Additional cost savings, through the efficient use
of the VM standby host’s memory resources for the VM guests, can be achieved by using the dynamic memory
allocation. Details of this functionality are described in the “dynamic memory allocation” section of this white paper.
Figure 8 shows an example of physical node consolidation within a single Serviceguard cluster by using multiple VM
guests running on a single physical node. This example could also be used in a single Serviceguard cluster “N+1”
configuration where a single VM guest serves as a standby node for the other nodes within a cluster.
Figure 8: Serviceguard cluster node consolidation using VM guests
In this example, a Serviceguard package on physical node 1 can failover to its VM guest 1 adoptive node running on
a VM host, while a package on the other VM guest 2 node running on the VM host can failover to its adoptive
physical node 2. This configuration provides cost savings by reducing the number of physical nodes required in the
cluster and by better system hardware utilization. This example also creates a configuration in which no more than
half of the cluster members are hosted on one physical server and alleviates the problem of the cluster losing quorum
due to one physical server failure (for example, the VM host with 2 cluster nodes).
It is also possible to consolidate multiple Serviceguard clusters using VM guests. Figure 9 shows an example of two
2-node Serviceguard clusters using four physical servers being consolidated onto two VM hosts. In this example, the
number of physical servers currently in-use is reduced by 50 percent, which results in the benefit of conserving data
center floor space and power usage.
Primary Node 1
Standby Host
VM Guest 1 VM Guest 2
Primary Node 2
Failover Failover
Serviceguard Cluster 2Serviceguard Cluster 1
Primary Node 1
Standby Host
VM Guest 1 VM Guest 2
Primary Node 2
Failover Failover
Serviceguard Cluster 2Serviceguard Cluster 1
Physical Node 1
VM Host
VM Guest 1 VM Guest 2
Physical Node 2
Failover
Failover
Serviceguard Cluster
Physical Node 1
VM Host
VM Guest 1 VM Guest 2
Physical Node 2
Failover
Failover
Serviceguard Cluster