Excessive memory consumption by SDN pods in ROSA

Solution Verified - Updated -

Environment

  • Red Hat OpenShift Service on AWS (ROSA)
    • 4

Issue

  • SDN pods are occupying too much memory.
  • Node is running out of memory resources due to high memory consumption by SDN pods.

Resolution

  • This issue is being tracked by the Engineering team in Red Hat Bugzilla 1370.
  • The workaround is to restart the SDN pods consuming too much memory.

$ oc delete pod sdn-xxxx -n openshift-sdn

Diagnostic Steps

  • Check the memory consumption of the nodes.

$ oc adm top nodes

NAME                                    CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%

ip-XX-X-XXX-XX.XXXXX.compute.internal    779m         9%     9684Mi          105%

ip-XX-X-XXX-XXX.XXXXX.compute.internal   340m         8%     3142Mi          100%
  • Check the memory consumption of the SDN pods.
NAME                 CPU(cores)   MEMORY(bytes)   

sdn-xxxx              900m         21143Mi         

sdn-xxxx              54m          22959Mi

This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.

Comments