View Issue Details

IDProjectCategoryView StatusLast Update
0014711BuildsysCi.centos.org Ecosystem Testingpublic2018-04-27 17:58
Reporterjlebon 
PrioritynormalSeverityminorReproducibilityhave not tried
Status resolvedResolutionfixed 
Summary0014711: Error sometimes when trying to `oc exec` into pods
DescriptionSometimes, when trying to `oc exec` into a pod, I get the following error:

Error from server: error dialing backend: dial tcp 172.22.6.92:10250: getsockopt: no route to host

Looking at https://github.com/openshift/origin/issues/18183, it seems like this can happen if port 10250 of the nodes are unreachable from the API server.
Other IPs for which I've seen this:

172.22.6.86:10250
172.22.6.89:10250
Steps To ReproduceProvision pods that get allocated on one of the problematic nodes. Try to `oc exec` into the pods.
TagsNo tags attached.

Activities

jlebon

jlebon

2018-04-23 21:32

reporter   ~0031668

This looks resolved now though it doesn't seem like I have the permissions to close my own bug.
Thanks for the quick action!
bstinson

bstinson

2018-04-27 17:58

administrator   ~0031688

Fixed node labels, and firewalld in apps.ci

Issue History

Date Modified Username Field Change
2018-04-23 18:13 jlebon New Issue
2018-04-23 21:32 jlebon Note Added: 0031668
2018-04-27 17:58 bstinson Note Added: 0031688
2018-04-27 17:58 bstinson Status new => resolved
2018-04-27 17:58 bstinson Resolution open => fixed