Troubleshooting monitoring

edit

Use the information in this section to troubleshoot common problems and find answers for frequently asked questions. See also Troubleshooting monitoring in Logstash.

For issues that you cannot fix yourself … we’re here to help. If you are an existing Elastic customer with a support contract, please create a ticket in the Elastic Support portal. Or post in the Elastic forum.

No monitoring data is visible in Kibana

edit

Symptoms: There is no information about your cluster on the Stack Monitoring page in Kibana.

Resolution: Check whether the appropriate indices exist on the monitoring cluster. For example, use the cat indices command to verify that there is a .monitoring-kibana* index for your Kibana monitoring data and a .monitoring-es* index for your Elasticsearch monitoring data. If you are collecting monitoring data by using Metricbeat the indices have -mb in their names. If the indices do not exist, review your configuration. For example, see Monitoring in a production environment.

Monitoring data for some Elastic Stack nodes or instances is missing from Kibana

edit

Symptoms: The Stack Monitoring page in Kibana does not show information for some nodes or instances in your cluster.

Resolution: Verify that the missing items have unique UUIDs. Each Elasticsearch node, Logstash node, Kibana instance, Beat instance, and APM Server is considered unique based on its persistent UUID, which is found in its path.data directory. Alternatively, you can find the UUIDs in the product logs at startup.

In some cases, you can also retrieve this information via APIs:

When you install Elasticsearch, Logstash, Kibana, APM Server, or Beats, their path.data directory should be non-existent or empty; do not copy this directory from other installations.