Generic CPU Usage (CPG1)

Description

Release notes

Show release notes
Version 1.1.0 - Early deployment
  • Improvement: Improve reliability when used on a physical hardware SOP (M13176)
  • Dependency:
    • SNMP Agent module version 3.4.0 or higher
    • Monitoring service >= 1.6.0 installed on the SMP (Contact support to find out the version currently installed)
    • SMP Base >= 1.13.0 installed on the SMP (Contact support to find out the version currently installed)

Version 1.0.0 - General deployment
  • Feature: Initial revision
  • Dependency:
    • SNMP Agent module version 3.4.0 or higher

Resource configuration interface

GUI unavailable.

What is measured

This probe returns a percentage of the CPU usage, for vSOP or physical SOP. Depending on the type of host, the measurement is done differently.

For the vSOPs

This probe returns a ratio that measures the time the vSOP and its processes spent in the cpu. It is expressed in % meaning the maximum value is 100. What is presented in the graph is the time that was spent between a probe inquiry and the next one.

The formula used to compute the ratio is:

(The actual time spent int the cpu's in seconds between 2 probe queries * 100) / (Number of CPU cores * time between 2 probe queries in seconds)

The number of CPU cores is found by the probe.

The time between 2 probes queries is 1 minute.

For the physical SOPs

This probe returns the usage percentage of the CPU as queried from SNMP. It is expressed in % meaning the maximum value is 100.

Resource parameters

  • Warning Level (default 80)
  • Critical Level (default 98)

Performance graphs

This probe generates a graph with following performance metrics:

  • CPU usage : The time spent in the CPU by the (v)SOP and its processes in seconds.
  • Warning threshold (optional): The time is in seconds and the threshold is fixed by the user.
  • Critical threshold (optional): The time is in seconds and the threshold is fixed by the user.

Alarms

This probe can report following alarm states:

  • WARNING: After 15 minutes during which the CPU usage in seconds surpassed the warning threshold.
  • CRITICAL: After 15 minutes during which the CPU usage in seconds surpassed the critical threshold.

Possible causes

  • The processes are requiring more processing time

Possible consequences

For a vSOP

The host's performance might be affected and other vSOP on the same host might be affected also because the vSOP is taking all the CPU resources for itself.

For a physical SOP

  • Degraded service like bad audio quality or slow applications
  • More difficult to get a remote console access

Possible actions

  • Connect to the SOP Shell in order to identify which process is in cause:
    • SOP Shell > Diagnostics > System > CPU/Memory usage
  • If it is the Communication Server, restart it by:
    • SOP Shell > Diagnostics > Communication Server > Open Console Then type "restart now"
  • If it is the Database Server, restart it by:
    • SOP Shell > Subsystems > MySQL Server > Restart"
  • Contact Escaux

Additional notes

In order to use the probe on a vSOP, it requires SNMP agent module to be installed on the VSOP itself.
Copyright © Escaux SA