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
Containerd >= 1.5.15 is affected by an issue that causes pod network info to not be written to disk, causing CNI info to be lost when containerd is restarted. This in turn causes the kubelet to restart all non-host-network pods whenever k3s is restarted.
Validated as part of patch validation on v1.23.16-rc2+k3s1
Restarting k3s agent did not restart the pods on fresh install or on upgrade from v1.23.15+k3s2 to v1.23.16-rc2+k3s1
backport of #6692
Containerd >= 1.5.15 is affected by an issue that causes pod network info to not be written to disk, causing CNI info to be lost when containerd is restarted. This in turn causes the kubelet to restart all non-host-network pods whenever k3s is restarted.
We need to update to v1.5.16-k3s2, which includes the backported fix from containerd/containerd#7845
The text was updated successfully, but these errors were encountered: