search

elastic / elastic-agent

Elastic Agent - single, unified way to add monitoring for logs, metrics, and other types of data to a host.
Other
126 stars 135 forks source link

Improve Elastic Agent integrations logging #380

Open mbudge opened 2 years ago

mbudge commented 2 years ago

We were affected by this bug when various Elastic Agent integrations would stop sending data yet the agent appeared healthy in the Fleet UI. There were no error/debug logs in the Elastic Agent logs to show the integration failed to start or the integration had stopped working.

https://github.com/elastic/beats/issues/30533

The state of the integrations could become difficult to monitor in a complex enterprise environment if the integration health isn't monitored as well as the agent health.

We're also finding it very time consuming having to log onto remote systems to collect Elastic Agent diagnostic logs, and remove all API and secret keys before proving them to support. On the prod network we don't have access to remote systems so a future improvement might be collecting Elastic Agent diagnostics through the Fleet interface in Kibana.

botelastic[bot] commented 2 years ago

This issue doesn't have a Team:<team> label.

jlind23 commented 2 years ago

@mbudge thanks for raising this, can I close this one as it seems to be a duplication of #100 ?

mbudge commented 2 years ago

Yes as long as we can monitor the state of the integrations from a single interface.

Also collecting elastic-agent diagnostic logs is very time consuming in an enterprise environment.

jlind23 commented 2 years ago

@nimarezainia do you know if we have any feature issue to collect agent diagnostic from fleet ui interface?