Navigation
This version of the documentation is archived and no longer supported. To learn how to upgrade your version of MongoDB Ops Manager, refer to the upgrade documentation.
You were redirected from a different version of the documentation. Click here to go back.

Fix Down Host

At time T, the Monitoring on host X had not detected a mongod process running on host Y port Z for A minutes.

Alert Conditions

You can configure alert conditions in the project-level alert settings page to trigger alerts.

To learn more about the alert condition, see Host is down.

Common Triggers

  • The mongod process could not be reached because of a network issue.
  • The mongod process is not running.
  • The host is powered off.

Fix the Immediate Problem

  • Verify network communication between the host running the Monitoring and the host that is listed as down. One way to do this is through mongosh:

    mongosh <hostname>:<port>
    

    If connecting from mongosh fails, try a basic ping test:

    ping <hostname>
    
  • Verify the mongod process is running:

    If the process was installed from deb or rpm package, query the service manager for the status:

    sudo service mongod status
    

    Alternatively, you can use ps:

    ps aux | grep mongo
    

Note

The Host is down alert should not be used to detect all occurrences of a host reboot. The alert is triggered only if the host is seen as down at the instant that the Monitoring runs the check. The interval between checks can vary from 1 to 3 minutes.

Implement a Long-Term Solution

Keep your host powered on and the mongod process running. Work to prevent any network issues.

Monitor Your Progress

View the following chart to monitor your progress:

  • Network

    Monitor network metrics to track network performance.

To learn more, see View Deployment Metrics.