Description of problem: Update migration-controller limits to value higher than 800Mi Customer reported needing at least 6Gi to stop the controller from being OOM Killed once they started using at scale Version-Release number of selected component (if applicable): 1.2.2 How reproducible: Always in a large enough environment Steps to Reproduce: 1. Deploy CAM 2. Creat a couple hundred plans 3. Actual results: Controller is OOM killed Expected results: controller is not OOM killed Additional info:
This possibly applies to the CPU limit as well.
Was addressed in PR https://github.com/konveyor/mig-operator/pull/410
Verified using MTC 1.3 openshift-migration-rhel7-operator@sha256:233af9517407e792bbb34c58558346f2424b8b0ab54be6f12f9f97513e391a6a Verified by running testcase OCP-34828 - Default controller limits value: Those are the new controller limits by default. DISCOVERY oc get deployment migration-controller -o "jsonpath={.spec.template.spec.containers[?(@.name=='discovery')].resources.limits}" map[cpu:1 memory:10Gi] CAM oc get deployment migration-controller -o "jsonpath={.spec.template.spec.containers[?(@.name=='cam')].resources.limits}" map[cpu:1 memory:10Gi] Moved to VERIFIED status.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Migration Toolkit for Containers (MTC) Tool image release advisory 1.3.0), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4148