Scale monitoring using metricbeat

[motkele (Moti Elkayam)]

Hello,

I have a very specific question regarding the way to collect metrics using metricbeat.
I would like to collect the metrics from all the services running on the server (and for which a metricbeat module is available).
I would like to better understand if Elastic provides some mechanism to perform an automatic discovery and config in that case:

• discover which services are running on the server on which the metricbeat agent is running
• configure elastic to start collecting metrics fron these services

Thanks,
Moti.

[ChrsMark (Chris Mark)]

Hi @motkele !

This is not supported in the generic case but can be achieved in specific environments like docker, kubernetes etc. You can check Autodiscover | Metricbeat Reference [7.16] | Elastic to have an idea of what is supported there.

C.

[motkele (Moti Elkayam)]

Hi Chris,
Thanks a lot of the prompt response.
What about VM based environments ?
I assume that each and every metricbeat user deploying on several dozens of servers is encountering the same challenge since he/she would not like to start and discover the services running on each server and then configure each one of them in a manual manner.
Moti.

[ChrsMark (Chris Mark)]

Yes I see the point here, hence the problem is too generic and there would be needed a mechanism to probe for known services' endpoints as well as a level of integration with each one of the orchestrators/hypervisors so as to automatically identify newcomer hosts.