Cluster power consumption efficiency performance monitor for HPC 2008 R2 cluster
This monitor tracks the power consumption efficiency of the compute nodes in the HPC cluster. The power consumption efficiency value is reflected by the utilization of the compute nodes, defined as the number of currently busy cores (the cores that have tasks allocated to them) divided by the number of currently available cores (the cores that are online and reachable).
This monitor is disabled by default. You can enable it and configure the power consumption threshold for Healthy and Warning states. (For information about how to override the threshold values, see the management pack guide.)
By default, the monitor is in a warning state when the power consumption efficiency of compute node cores in the cluster is less than 40% at five consecutive sampling points (it samples every 5 mins while the frequency is configurable). At a higher percentage, the monitor is in a Healthy state.
You can enable one of the power consumption adjustment rules to increase the cluster power consumption efficiency:
Calendar-based: Changes the power state (hibernating or powered on) of a percentage of the compute nodes during specific times each day.
Utilization-based: Changes the power state (hibernating or powered on) of inactive compute nodes
Target | Microsoft.HPC.2008R2.ActiveHeadNode | ||
Parent Monitor | System.Health.PerformanceState | ||
Category | PerformanceHealth | ||
Enabled | False | ||
Alert Generate | True | ||
Alert Severity | MatchMonitorHealth | ||
Alert Priority | Normal | ||
Alert Auto Resolve | True | ||
Monitor Type | Microsoft.HPC.2008R2.MonitorType.PowerConsumptionMonitor | ||
Remotable | True | ||
Accessibility | Public | ||
Alert Message |
| ||
RunAs | Microsoft.HPC.RunAsProfile.AdminActionAccount |
<UnitMonitor ID="Microsoft.HPC.2008R2.Monitor.HeadNode.Performance.PowerConsumption" Accessibility="Public" Enabled="false" Target="Microsoft.HPC.2008R2.ActiveHeadNode" ParentMonitorID="Health!System.Health.PerformanceState" Remotable="true" Priority="Normal" RunAs="HPCLibrary!Microsoft.HPC.RunAsProfile.AdminActionAccount" TypeID="Microsoft.HPC.2008R2.MonitorType.PowerConsumptionMonitor" ConfirmDelivery="true">
<Category>PerformanceHealth</Category>
<AlertSettings AlertMessage="Microsoft.HPC.2008R2.Monitor.HeadNode.Performance.PowerConsumption_AlertMessageResourceID">
<AlertOnState>Warning</AlertOnState>
<AutoResolve>true</AutoResolve>
<AlertPriority>Normal</AlertPriority>
<AlertSeverity>MatchMonitorHealth</AlertSeverity>
</AlertSettings>
<OperationalStates>
<OperationalState ID="UIGeneratedOpStateId2fab8a3b60b34522aef5c20d9ca4a347" MonitorTypeStateID="Low" HealthState="Warning"/>
<OperationalState ID="UIGeneratedOpStateId89b8bd8ca6fc41ccbddbfce327846b40" MonitorTypeStateID="High" HealthState="Success"/>
</OperationalStates>
<Configuration>
<ClusterName>$Target/Property[Type="Microsoft.HPC.2008R2.ActiveHeadNode"]/ClusterName$</ClusterName>
<Threshold>40</Threshold>
<NumSamples>5</NumSamples>
<TimeoutSeconds>300</TimeoutSeconds>
<IntervalSeconds>300</IntervalSeconds>
</Configuration>
</UnitMonitor>