Managing HP Serviceguard A.12.00.00 for Linux, June 2014

9 Cluster Analytics
Serviceguard cluster analytics provides a mechanism to the users to perform "availability audits"
on Serviceguard cluster, nodes, and application packages running on the clusters.
Key Performance Indicators and Key Events of Cluster Analytics
Key Performance Indicators, also known as KPIs, is a metric of Serviceguard cluster object, such
as cluster, node, or package for which a value can be obtained from the cluster analytics engine
for a specified time period. For example, a package availability value can be computed for a time
period selected by the user. For more information, see cmcashowkpi (1m) manpage.
Key Events, also known as KEs, in Serviceguard cluster is an event which is important from the
perspective of availability and change management. Key events are important because the [date
stamp, key event] pair allows you to reconstruct the entire history of any monitored object in the
cluster. The availability of a monitored entity is computed by looking at the key events of the entity
and the corresponding date stamps.
Supported KPIs of Cluster Analytics
Table 15 describes the supported KPIs of the cluster, nodes, and packages:
NOTE: You can obtain better visualizations for KPIs using GUI.
Table 15 KPIs of the Cluster, Nodes, and Packages
KPIs of the Cluster
The percentage of time duration for which cluster is up for the
total time specified. By default, the total time is considered to be
the difference between current time and time at which cluster is
created. You can also specify the start and end time for which
the KPI value will be computed.
AVAILABILITY
The percentage of time duration for which cluster was brought
down gracefully for the total time specified. By default, the total
time is considered to be the difference between current time and
time at which cluster is created. You can also specify the start
and end time for which the KPI value will be computed.
MAINTENANCE
The time when the cluster was last reformed.LAST_REFORMATION_TIME
The time when the cluster was created.CREATION_TIME
The number of times a cluster is reformed.
1
REFORMATION_COUNT (This can be an indicator of
cluster stability)
The time when the cluster was last modified.LAST_MODIFICATION_TIME
This is the ratio of available nodes to the total number of nodes
configured in a cluster.
FAILURE_PROTECTION_LEVEL
KPIs of Nodes in the Cluster
The percentage of time duration for which node is up in the cluster
for the total time specified. By default, the total time is considered
to be the difference between current time and time at which node
is added to the cluster. You can also specify the start and end
time for which the KPI value will be computed.
AVAILABILITY
The percentage of time duration for which node was brought
down gracefully for the total time specified. By default, the total
time is considered to be the difference between current time and
time at which node is added to the cluster. You can also specify
the start and end time for which the KPI value will be computed.
MAINTENANCE
263