Check helper pod status is failing because of multiple same chaos experiments running in the k8s namespace about litmus-go HOT 6 CLOSED

Thank you @songrgg for reporting this. We will get back on this soon.

from litmus-go.

This is fixed via PR #276 & available since 1.12.2

from litmus-go.

After I checked the code, it seems it's using a generic appLabel app=app=pod-network-loss-helper to filter the helper pods, which will be mixed by different application's chaos experiment pods. Do I understand correctly?

from litmus-go.

@songrgg you are right. We are in fact using a RunID in order to add a dynamic element to the helper pods to differentiate it from the different instances of the helper belonging to other experiments, but filtering by a common label leading to this problem.

This will be addressed by changing the choice of label used in filtering the right helpers to get their statuses. This will be made available in 1.12.2 which should be available by 22/01 (along with few more useful changes).

from litmus-go.

Sounds great, thank you :)

from litmus-go.

I think ChaosUID might be a good option.

from litmus-go.