[prod-preview] After resetting environment che pod fails to start - "Failed create pod sandbox"

[ibuziuk]

Che 2.0.67 on https://console.free-int.openshift.com

Steps to reprodue: Profile -> Update Profile -> Reset Environment

Events:

[vpavlin]

I've seen this with Jenkins yesterday whole day. It came up after some time. I found this openshift/origin#17047

[ibuziuk]

Currently this issue is a blocker for testing che on prod-preview

[kbsingh]

@vpavlin can we get a seperate issue filed for jenkins

[ibuziuk]

Still can reproduce on prod-preview:

[ibuziuk]

@kbsingh After a few attempts I was able to deploy Che on console.free-int.openshift.com, so probably we should change the severity to SEV2-high

[vpavlin]

Yeah, same for Jenkins - this is not 100 % reproducible

[joshuawilson]

@kbsingh do you want me to add a label area/jenkins that we can filter by? Most of these issues get filed under area/pipelines.

[kbsingh]

yes please, Also - talk in the SD team, we are ok to drop SEV1 to SEV2 if the owning dev is happy with that.

@joshuawilson @ibuziuk

[xyntrix]

Likely related to an issue with re-pinning of free-int/free-stg, where data is in a de-sync'd state.

#1550 (comment)

[xyntrix]

#1550 (comment) is now resolved. can we get this re-tested and if it's all great, let's close this out.

[xyntrix]

correction, the pod sandbox issue may be related to aws limits here: #1538 -- asking for some validation around this.

[ibuziuk]

@mmclanerh can not reproduce anymore but for some reason there are only 3 projects on free-int for my account:

[kbsingh]

@mmclanerh i dont think the sandbox issues are related to aws limits; note that the clustwr with the limit problem does not have sandbox issues.

[kbsingh]

@ibuziuk i would suggest chasing the projects issue on a seperate issue. Closing sandbox problem for now.

[ibuziuk]

@kbsingh separate issue for missing namespaces - #1612