Go through the Kubelet logs and application pod logs to know the reason for marking the ReadOnly and take appropriate action. cStor target has lost quorum because of underlying node losses and target has marked the lun as ReadOnly.During this time, the iSCSI initiator at the application pod has timeout and marked the underlying filesystem as ReadOnly Kubernetes can take up to 30 minutes as timeout before deciding the node is going to stay offline and pods need to be rescheduled. Node is rebooted in adhoc manner (or unscheduled reboot) and Kubernetes is waiting for Kubelet to respond to know if the node is rebooted and the pods on that node need to be rescheduled.The cStor target pod is evicted because of resource constraints and is not scheduled within time.Persistent volumes indefinitely remain in pending state Application complaining ReadOnly filesystem #Īpplication sometimes complain about the underlying filesystem has become ReadOnly. Unable to provision cStor on DigitalOcean Unable to mount XFS formatted volumes into Pod Unable to clone OpenEBS volume from snapshot Recovery procedure for Read-only volume for XFS formatted volumes Recovery procedure for Read-only volume where kubelet is running in a container
OpenEBS Jiva PVC is not provisioning in 0.8.0
#Globalsan iscsi under recovery 10.12 verification#
Unable to create persistentVolumeClaim due to certificate verification errorĪpplication pods are not running when OpenEBS volumes are provisioned on RancherĪpplication pod is stuck in ContainerCreating state after deploymentĬreating cStor pool fails on CentOS when there are partitions on the diskĪpplication pod enters CrashLoopBackOff state Search for any reported issues on StackOverflow under OpenEBS tagĪpplication complaining ReadOnly filesystem.Search for similar issues added in this troubleshooting section.
0 kommentar(er)
