Hi all
At the last few Updates I always got troubles on one Node. The docker contrainer was pulled, the old was deleted but the new could not automatically started. I need to start the docker then once manually. After I started id manually everything works well. My Node is now about 11 months old and I got these troubles only since about the last 3 updates. Is there something I can do? The Problem is that when my node is updating at the night and I only see it in the morning. The offline time would be very close or are larger then allowed. I don’t want to be disqualifed in the future for this.
Hey, for now we are working on auto-update logic for docker distribution, including ability to restart docker container after update. We hope these changes will help to avoid this situation.
I would not worry about DQ right now as DQ for downtime is still disabled. But this is certainly a problem you’ll want to get to the bottom of. I would suggest removing and reinstalling watchtower. And just to be sure we should remove old images.
Make sure there aren’t any instances of watchtower running, possibly under a different name. Then:
docker system prune -a
This will remove old images from the cache. Then reinstall watchtower as per the official docs: https://documentation.storj.io/setup/cli/software-updates I think you can add the --debug flag to the watchtower run command to get a bit more in the watchtower logs.
You can leave the storagenode container running. docker system prune only removes old images and current images for non-running containers. So really, it is best to leave the storagenode containers running. More info:
Okay, found another watchtower docker is running… Should I stop and delete this?
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
166b8a8fc759 storjlabs/storagenode:latest "/entrypoint" 34 hours ago Up 34 hours 0.0.0.0:14002->14002/tcp, 0.0.0.0:28967->28967/tcp storagenode
5a20d45c9076 84803293c0e3 "/watchtower storage…" 11 months ago Up 35 hours eXrNzqZkfIQnkbTDzmgxbtgezXkcGNmk