Data gathered for performance on Linux®

The following data is obtained by the gpfs.snap command, if the option --performance is provided.

  • The output of the command top -n 1 -b on all nodes.
  • The performance metrics data gathered on the node having the ACTIVE THRESHOLD MONITOR role. The output contains the metrics data of the last 24 hours, collected from all the metrics sensors enabled on the cluster.
    Note: If the frequency of a sensor is greater than 60 seconds, then the data gathering is performed with the original frequency of the metrics group. If the frequency of a sensor is less than or equal to 60 seconds, the frequency is set to 60 seconds.
  • The performance data package includes the performance monitoring report from the top metric for last 24 hours as well.

For gathering performance data, it is recommended to execute the gpfs.snap command with the -a option on all nodes, or at least on the pmcollector node having the ACTIVE THRESHOLD MONITOR role. For more information about the ACTIVE THRESHOLD MONITOR role, see Active threshold monitor role.