Why are Openshift Container Storage Nodes Observing 'Disk Pressure' Events When a Soseport is Being Collected?
Issue
-
Nodes hosting Gluster pods in Openshift Container Storage, are facing
Disk pressure
events, leading to pod evictions:15s 15s 1 glusterfs-pod1 Pod Warning Evicted kubelet, node1.example.cm The node had condition: [DiskPressure]. 14s 14s 1 glusterfs-pod1 Pod Warning Evicted kubelet, node1.example.com The node had condition: [DiskPressure].
-
At the same time, a
sosreport
with the--all-logs
option was being collected. -
Why is the node facing disk pressure for the collection of a sosreport? How to avoid these events?
Environment
Red Hat Openshift Container Storage version 3.x, configured in converged mode.
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.