Automatic Dynatrace deployment on Kubernetes - verification of agent

[mariusz_cwiklin]

Our deployment of Dynatrace monitoring on Kubernetes is automated and managed by different team.

We have seen several cases where dynatrace-operator and dynakube-classic pods have been running but no data was sent to the Dynatrace cluster.

We opened a case to Dynatrace support and they said it was because of bug (This was a known issue and caused the agent being corrupted while pulling OA 1.223 by API via AG v221.).

So we need a way to verify the status of Dynatrace deployment in case when some bugs will block Dynatrace agent. Support told that the only way is to check manually the Dynatrace UI console to see if new agent has appeared, which is not a good solution for automated process.

Is there a way to check that Dynatrace agent on Kubernetes worker node (dynakube-classic) has been successfully deployed and connected to Dynatrace cluster?

[ChadTurner · Accepted solution]

as a general rule for us we are always Minus 1 version from the latest version. All new oneagent deployments are also tested on a group of hosts that have representative for each type of service offering, cloud, vmware, onprem etc.... 

There is no automation to test this from what im aware of, hence our manual deployments and tests. 

You could go through the UI and check the versions from the deployment Status page and set the MZ of your targeted upgrade, but no automation methods. 

[mariusz_cwiklin]

This is very bad that in the era of DevOps/GitOps to verify Dynatrace successful deployment on Kubernetes you have to manually go to Dynatrace console and check the result.