Too many haproxy processes running in default-router pod on RHOCP node
Issue
- High number of
haproxy
processes indefault-router
pod
causinghigh load average
onRHOCP
node. - Ingress node memory consumption is growing until all memory is consumed, causing OOM (out-of-memory) killing of random processes and overall ingress node outage.
- Router pods are intermittently restarting without explanation a few times every day/hours.
Environment
- Red Hat OpenShift Container Platform (RHOCP) 4.x
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.