URL:
Reporter RHNID:
Section: -
Language: en-US (English)
Workaround: Be sure students are deleting old projects (helps but doesn't solve it). on master: systemctl restart atomic-openshift-* when it happens.

Description: First time running v3.9 DO280.  This was on the 32G, core i7, ssd laptops.

Several students, and myself, have experienced bad delays in executing commands such as oc get pods and ssh root@master.  Wait long enough and it would eventually respond; but the wait could be from several seconds to over a minute.

Investigating, I noted high cpu usage on our foundations from time to time - eventually narrowed it master taking up the most cpu cycles.  I also noted low available memory.  Running top, it appears the culprit is kswapd0.

I believe this can be resolved by allocating more memory to master.  We have it available - DO280 is a Level IV class now so...