Implement e2e test that kills KEDA Operator & Metrics Server pods #3927
Labels
feature-request
All issues for new features that have not been committed to
help wanted
Looking for support from community
needs-discussion
testing
Proposal
Implement e2e test that kills KEDA Operator & Metrics Server pods while scaling is happening, the test should check that KEDA is able to continue in the scaling after errors on both.
Scenarios:
We should run all these scenarios while having 1 replica of Operator and MS and also when we have mutliple replicas of these.
Use-Case
No response
Anything else?
No response
The text was updated successfully, but these errors were encountered: