che kubernetes-image-puller-operator 1.0.0 is in the CrashLoopBackOff State due to low memory limits

Describe the bug

Che version

other (please specify in additional context)

Steps to reproduce

Install 1.0.0 image puller on a cluster with 100+ namespaces ERROR: Operator pod is crashing

Expected behavior

Operator pod is running without errors

Runtime

OpenShift

Screenshots

No response

Installation method

OperatorHub

Environment

Amazon

Eclipse Che Logs

No response

Additional context

Probably the limits are low:

Resource requests
cpu: 100m, memory: 64Mi
Resource limits
cpu: 100m, memory: 128Mi

Jun 08 '22 13:06 ibuziuk

Should be fixed in the 1.0.1 upcoming release

Jun 23 '22 09:06 ibuziuk

looks like current limits are not enough for large clusters. Installation failed on the developer sandbox. As a workaround the memory limit in csv has been updated to 1Gi

Aug 01 '22 16:08 ibuziuk

@dkwon17 I believe we need to release 1.0.2 with more generous limits for operator

Aug 01 '22 16:08 ibuziuk

According to the metrics we probably need to set it to 1Gi to be on the safe side. During the startup there is a spike, but then RAM consumption drops to ~ 400M

Aug 02 '22 10:08 ibuziuk

@ibuziuk I made a PR: https://github.com/che-incubator/kubernetes-image-puller-operator/pull/83

Aug 02 '22 18:08 dkwon17

Fixed in 1.0.2 version https://operatorhub.io/operator/kubernetes-imagepuller-operator Closing

Aug 16 '22 11:08 ibuziuk