Description of problem:
governance-policy-framework-addon is passing memory limits on managed cluster
Version-Release number of selected component (if applicable):
2.7.4, 2.8
How reproducible:
situational
Steps to Reproduce:
- set up a high numer of policies (512 here)
- apply them on managed clusters
Actual results:
memory limits are hit and the pods get OOMkilled then restarted and enter a loop of getting oomkilled
Expected results:
no oomkilled loop
Additional info:
512 policies are in place.
the issue was present in 2.7 and is in 2.8. raising the limit with https://access.redhat.com/solutions/7020060 works around the issue but the memory limit here was raised to 2GiB to work around with 2.8.
customer currently on 2.8.0 at time of writing
- is cloned by
-
ACM-6783 [2.8] governance-policy-framework-addon getting OOMkilled on managed clusters
-
- Closed
-
