Workload metrics for Openshift - Limitation when process group detection rules are required

[Michael_Brunell]

Hello,

I am facing an issue that is preventing me from enabling the Show workloads and cloud applications toggle to capture Openshift metrics. Once enabled, all the process group detection rules that I have created would be ignored

Since we're using blue/green pods deployments, I need to merge PGs together. Otherwise, data is splitted over 2 different services into 2 different process groups as explained into this post

https://community.dynatrace.com/spaces/482/dynatrace-open-qa/questions/248228/is-it-possible-to-see-...

Here is an example of the existing PG detection rules...

These PG rules allow me to group the blue and green containers into 1 single PG! and into the same service!

Any idea of what can be done to allow me to capture the Openshift metrics without splitting the blue/green data into 2 different services? Any workaround?

Thanks!

[ChadTurner · Accepted solution]

this is a really clever workaround. Other then merging services and such, id recommend tossing in a RFE and referencing this post as well as the other and maybe Dynatrace can build functionality into the product for customers that have a similar set up

[Michael_Novich]

i'm leaning toward blue-green and have run into the problem as well - i'd have to turn off the workloads detection through kubernetes integration to be able to have them as part of a single process group and then merge all the services. i'd lose all the existing history and rules we have in place for request renaming etc. Really wish there was an easier way but the cloud workloads enforce grouping that cant be overridden it seems.

[melorymonie]

Hi Michel,

After creating those PG Rule you have to Merge both services?

I'm in the same situation but I couldn't merge the services as blue/green pods never run at the same time.

is there any solution for this case? Any RFE?

Regards