topic Re: When OneAgent itself crashed, how we know that? in Open Q&A

When OneAgent itself crashed, how we know that?

nkobayashi1 — Wed, 20 Jun 2018 09:45:55 GMT

Hi,

When the OneAgent crashed and the monitoring stopped, how can we know that?
We have been requested that by our customer.
Any information would greatly help.

Thanks,
Noah Kobayashi

Re: When OneAgent itself crashed, how we know that?

Julius_Loman — Wed, 20 Jun 2018 10:40:33 GMT

Oneagent has a watchdog process, that will start up any crashed oneagent subprocesses (e.g. oneagentos, oneagentplugin... ). The deep-instrumented processes will continue to operate and send data even if Oneagent is shut down, since the instrumentation and data collection is not dependent on Oneagent running.

Re: When OneAgent itself crashed, how we know that?

pahofmann — Wed, 20 Jun 2018 14:19:42 GMT

As Julius mentioned, the watchdog will automatically restart it.

If you wan't to know if it crashed you can check the Event Viewer on Windows or check the service status with systemd on linux.

Re: When OneAgent itself crashed, how we know that?

alvin — Sun, 21 Mar 2021 05:16:50 GMT

Julius,

Do you mean to say deep-instrumentation data collection will continue to happen even if OneAgent has gone down but sending of this data to the Dynatrace server will only happen once OneAgent comes back up? How will Dynatrace server receive the data if not through OneAgent, is there another way? Thanks for clarifying.