You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 3, 2021. It is now read-only.
The validated docker versions are the same as for v1.10: 1.11.2 to 1.13.1 and 17.03.x (ref)
The reason we noticed this problem is that we're seeing issues where secrets volumes fail to unmount when cleaning up a large number of pods. Once the volumes get locked up failing to unmount, the pods all get stuck in a Terminating state that can't be resolved until someone manually restarts the Docker daemon (at which point they'll clear out unless they hit the same lock problem). This issue indicates that this may well be a problem with the version of Docker we're deploying.
What happened:
The version of Docker running in a cluster is
17.12.1-ce
:What you expected to happen:
I would expect CFCR to ship with one of the officially validated versions of Docker:
From Changelog v1.10:
How to reproduce it (as minimally and precisely as possible):
Deploy a cluster with CFCR v0.19,
ssh
to a worker, and run the command seen above.Anything else we need to know?:
The Docker versions are the same for Kubernetes 1.11:
The reason we noticed this problem is that we're seeing issues where secrets volumes fail to unmount when cleaning up a large number of pods. Once the volumes get locked up failing to unmount, the pods all get stuck in a
Terminating
state that can't be resolved until someone manually restarts the Docker daemon (at which point they'll clear out unless they hit the same lock problem). This issue indicates that this may well be a problem with the version of Docker we're deploying.For comparison, GKE is only shipping with Docker v17.03 as well:
The text was updated successfully, but these errors were encountered: