-
Notifications
You must be signed in to change notification settings - Fork 1.1k
Description
Description
This is random issue that has been happening on kata-containers CI, with the "VM" runtime type, for a while.
Please, refer to: http://jenkins.katacontainers.io/job/kata-containers-2-crio-PR/16/console
16:59:13 # status:
16:59:13 # conditions:
16:59:13 # - lastProbeTime: null
16:59:13 # lastTransitionTime: "2021-05-17T14:57:43Z"
16:59:13 # status: "True"
16:59:13 # type: Initialized
16:59:13 # - lastProbeTime: null
16:59:13 # lastTransitionTime: "2021-05-17T14:57:43Z"
16:59:13 # message: 'containers with unready status: [oom-test]'
16:59:13 # reason: ContainersNotReady
16:59:13 # status: "False"
16:59:13 # type: Ready
16:59:13 # - lastProbeTime: null
16:59:13 # lastTransitionTime: "2021-05-17T14:57:43Z"
16:59:13 # message: 'containers with unready status: [oom-test]'
16:59:13 # reason: ContainersNotReady
16:59:13 # status: "False"
16:59:13 # type: ContainersReady
16:59:13 # - lastProbeTime: null
16:59:13 # lastTransitionTime: "2021-05-17T14:57:43Z"
16:59:13 # status: "True"
16:59:13 # type: PodScheduled
I thought we were able to track it down and solve it with #4803 and #4879 but the issue still happens every now and then.
Locally, so far, I wasn't able to reproduce it with a few hundred runs, but I'll once I am able to reproduce it I'll add the steps to reproduce the issue.
Steps to reproduce the issue:
1.
2.
3.
Describe the results you received:
Containers never reach the "Created" status.
Describe the results you expected:
Containers would get their status to Created
Additional information you deem important (e.g. issue happens only occasionally):
This is hard to reproduce, as it happens randomly. /o\
Output of crio --version:
It's happening with master and release-1.21 branches, and the issue was originally introduced as part of #4650
Additional environment details (AWS, VirtualBox, physical, etc.):