On killing the processes that were still running
after ungraceful restart, containerd can’t fully
exit the process if the fifos have not been fully
read. In this case process was just marked as exited
and not removed from the internal containers array.
Fixes#22913
Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
Restore the 1.10 logic that will reset the restart manager's timeout or
backoff delay if a container executes longer than 10s reguardless of
exit status or policy.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
(cherry picked from commit b6db56b5eb)
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
This avoid an extra bind mount within /var/run/docker/libcontainerd
This should resolve situations where a container having the host
/var/run bound prevents other containers from being cleanly removed
(e.g. #21969).
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
(cherry picked from commit 3135874543)
It should not be an error to call a common option to cancel restarts.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
(cherry picked from commit fc2e2234c6)
Don't throw "restartmanager canceled" error for no restart policy container
and add the container id to the warning message if a container has restart policy
and has been canceled.
Signed-off-by: Lei Jitang <leijitang@huawei.com>
(cherry picked from commit 494297baf8)
When user try to restart a restarting container, docker client report
error: "container is already active", and container will be stopped
instead be restarted which is seriously wrong.
What's more critical is that when user try to start this container
again, it will always fail.
This error can also be reproduced with a `docker stop`+`docker start`.
And this commit will fix the bug.
Signed-off-by: Zhang Wei <zhangwei555@huawei.com>
(cherry picked from commit a705e166cf)
