fix: filter out deamon pod and unnecessary containers from metrics output

[adarsh0728]

1. Filter out daemon pod from metrics response.
2. Filter in only pod's containers from metrics response.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.02%. Comparing base (7d2251d) to head (e934f1f).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2394      +/-   ##
==========================================
- Coverage   71.10%   71.02%   -0.08%     
==========================================
  Files         367      367              
  Lines       53042    53042              
==========================================
- Hits        37715    37674      -41     
- Misses      14247    14293      +46     
+ Partials     1080     1075       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[adarsh0728]

[vigith]

@adarsh0728 what is the condition to filter out pods and containers? keep in mind that in future we will support builtins like kafka, sqs, etc., we cannot filter those out.

[adarsh0728]

@adarsh0728 what is the condition to filter out pods and containers? keep in mind that in future we will support builtins like kafka, sqs, etc., we cannot filter those out.

For pods: we call pods API and filter out only daemon.
For containers of a pod: pod.spec.containers + pod.spec.initContainers(those with restartPolicy as Always)

we are not filtering out builtins with this

[veds-g]

for filtering the containers, we do not have the pipeline or monovertex context is the metrics response. So in case of multiple containers with the same name, how do we ensure that we showing data for the particular container belonging to that pipeline vertex or monovertex?

[veds-g]

LGTM