Integration with Amazon CloudWatch Logs - AWS ParallelCluster

Integration with Amazon CloudWatch Logs

Starting with AWS ParallelCluster version 2.6.0, common logs are stored in CloudWatch Logs by default. For more information about CloudWatch Logs, see Amazon CloudWatch Logs User Guide. To configure CloudWatch Logs integration, see the [cw_log] section and the cw_log_settings setting.

A log group is created for each cluster with a name /aws/parallelcluster/cluster-name (for example, /aws/parallelcluster/testCluster). Each log (or set of logs if the path contains a *) on each node has a log stream named {hostname}.{instance_id}.{logIdentifier}. (For example ip-172-31-10-46.i-02587cf29cc3048f3.nodewatcher.) Log data is sent to CloudWatch by the CloudWatch agent, which runs as root on all cluster instances.

Starting with AWS ParallelCluster version 2.10.0, an Amazon CloudWatch dashboard is created when the cluster is created. This dashboard makes it easy to review the logs stored in CloudWatch Logs. For more information, see Amazon CloudWatch dashboard.

This list contains the path of the logs and the logIdentifier used for those logs.

  • /opt/sge/default/spool/qmaster/messages (sge-qmaster)

  • /var/log/cfn-init.log (cfn-init)

  • /var/log/chef-client.log (chef-client)

  • /var/log/cloud-init.log (cloud-init)

  • /var/log/cloud-init-output.log (cloud-init-output)

  • /var/log/dcv/agent.*.log (dcv-agent)

  • /var/log/dcv/dcv-xsession.*.log (dcv-xsession)

  • /var/log/dcv/server.log (dcv-server)

  • /var/log/dcv/sessionlauncher.log (dcv-session-launcher)

  • /var/log/dcv/Xdcv.*.log (Xdcv)

  • /var/log/jobwatcher (jobwatcher)

  • /var/log/messages (system-messages)

  • /var/log/nodewatcher (nodewatcher)

  • /var/log/parallelcluster/clustermgtd (clustermgtd)

  • /var/log/parallelcluster/computemgtd (computemgtd)

  • /var/log/parallelcluster/pcluster_dcv_authenticator.log (dcv-authenticator)

  • /var/log/parallelcluster/pcluster_dcv_connect.log (dcv-ext-authenticator)

  • /var/log/parallelcluster/slurm_resume.log (slurm_resume)

  • /var/log/parallelcluster/slurm_suspend.log (slurm_suspend)

  • /var/log/slurmctld.log (slurmctld)

  • /var/log/slurmd.log (slurmd)

  • /var/log/sqswatcher (sqswatcher)

  • /var/log/supervisord.log (supervisord)

  • /var/log/syslog (syslog)

  • Note

    Starting with version 2.11.5, AWS ParallelCluster doesn't support the use of SGE or Torque schedulers.

    /var/spool/sge/*/messages (sge-exec-daemon)

  • /var/spool/torque/client_logs/* (torque-client)

  • /var/spool/torque/server_logs/* (torque-server)

Jobs in clusters that use AWS Batch store the output of jobs that reached a RUNNING, SUCCEEDED, or FAILED state in CloudWatch Logs. The log group is /aws/batch/job, and the log stream name format is jobDefinitionName/default/ecs_task_id. By default, these logs are set to never expire, but you can modify the retention period. For more information, see Change log data retention in CloudWatch Logs in the Amazon CloudWatch Logs User Guide.

Note

chef-client, cloud-init-output, clustermgtd, computemgtd, slurm_resume, and slurm_suspend were added in AWS ParallelCluster version 2.9.0. For AWS ParallelCluster version 2.6.0, /var/log/cfn-init-cmd.log (cfn-init-cmd) and /var/log/cfn-wire.log (cfn-wire) were also stored in CloudWatch Logs.