Dynatrace API restart status for Process Group Instance

[c_m_anderson]

I've combed through the API data quite a bit for Process Groups, Processes and Services, searching for a property to help us identify through the API when Dynatrace requires a process to be restarted in order to continue monitoring.

I'm basically trying to identify the state in the following photo via API.

[wolfgang_beer · Accepted solution]

There is an enhancement on the way already that allows you to query the monitoring state of processes through the normal processes API. I think that will help you to solve your use-case. This enhancement should be available in Dynatrace SaaS within the next 2 weeks.

[c_m_anderson]

This is great news. Thanks Wolfgang!

[israel_ochoa]

Hi @Wolfgang B. is this feature available now? thanks.

[wolfgang_beer]

It should for Dynatrace SaaS deployments. Dynatrace Managed should come with release 150. You can easily check if its available by fetching the process group API which returns the agent injections state and if a restart is necessary.

[israel_ochoa]

Thanks @Wolfgang B.

[apm-bradesco1]

Hi @Wolfgang B.,

Is this feature available now? We are on Version 1.154.146.20181010-132746 of Dynatrace Managed.

I did some API tests but I couldn't find this feature.

Thanks!

[wolfgang_beer]

You find the agent status only on the /api/v1/infrastructure/processes endpoint. Did you check that one?

[apm-bradesco1]

It worked! I was finding for /api/v1/infrastructure/process-group.

Thanks!

[wolfgang_beer]

Excellent 🙂

[caio_marfil]

Hi

@Wolfgang B. in our test enviroment(3 Dynatrace Nodes) we can get the JSON after some time outs, in Production enviroment(10 Dynatrace Nodes) we can not get the API JSON due to the time out, even after 100 tries.

Can we extend the time out time for production enviroment? Now, it seems to be configured with the default of 30 seconds.

Thanks!

[wolfgang_beer]

Please try to filter the result set by using tags or management zones. The overall result set is simply to huge to fetch it in one query. We are working on speeding up things right now but also in the future you will have to filter huge environments before you get a reasonable fetchable result.

[caio_marfil]

Understood... I also would like to understand why some times we get time out and some times the request responds in 2 seconds(the same request in Test enviroment)...

Is there any queue for this requests..?

[caio_marfil]

and do you know the paramter to filter the management zone? I tried many ways and I couldn't find anywherere.

[carlos_henrique]

Hello @Wolfgang B.! My question is the same of @Israel O. is this feature available now for managed environment? If yes, Could you provider an example? Thank you so much.

[wolfgang_beer]

Yes it is. Use the processes API as below:

/api/v1/entity/infrastructure/processes

And check for the monitoring state attribute, as shown below:

[carlos_henrique]

Perfect @Wolfgang B.! Thank you so much!