after the most recent storj update with watchtower, my docker froze. On reboot It would say docker was starting but would never fully start and just freeze. Docker was updated to latest version, reboot and docker still froze and wouldn’t finish starting. Only way around was I had to uninstall docker then reinstall and then docker came back to life.
Next problem I had to pull the new storagenode:beta but the message is no matching manifest amd64/windows 10.0.18362. per documentation setup instructions.
running windows 10 pro. didn’t have a problem with the last automatic update and everything was running fine until this morning at 8AM CT. looks like storagenode quit to do an update and froze.
If your run command is exactly as shown in the support docs then you shouldn’t face this problem. I wanted to make sure you had the right format of the command but I trust you do.
The docker pull command precedes the run command. So @eagleye is right that the run command is irrelevant at that point. I’m not sure what’s going on why this doesn’t work though.
Not if a local version of that image already exists. This will just use the old image and not pull the updated one. I’m also pretty sure that if you do purge dangling images first, the run command would run into the same issue as the pull command.
That’s not what happened with me. I clearly remember I did prune, removed all containers, reinstalled docker then just executed the docker run and it pulled the required image. I had those =======> progress bars too for downloading new image.
In windows? I always had them as windows containers prior to these updates. Any special commands on docker for windows after I do that?
Ok That worked.
Uninstalling and reinstalling docker was needed because prior to this problem docker wasn’t running well. I saw that switch but it also kept hanging until I uninstalled docker.
Yes. The storagenode is a Linux image. So, you must have a Linux containers enabled in Docker desktop for Windows.
Right click on the Docker desktop icon in the system tray and switch back to the Linux containers. This is default setting for Docker desktop.
I don’t know, why you have switched to the Windows containers.
The hanging probably related to
I couldn’t see the watchtower hang in windows as obvious as in Linux. But, uninstalling and reinstalling docker disabled my watchtower and I will suspend running watchtower until I am sure the watchtower issue is fixed. Yes, my problems occurred as watchtower executed an update and my docker hung. I saw docker activity in processes running abnormally, but I didn’t know what was normal or abnormal at the time.
A followup, After the update and fix and everything I did above, my node dashboard is only reporting 300GB of space used, yet prior to these problems I had over 1.6TB of data stored. My audits are running and not showing errors. Checking the folder and I still have 1.6TB of data stored in the node.
I ran the database repair command and it came back OK.
I did receive an error message after the storagenode update to 22.1 that the database was splitting and then there was an error. My log file no longer has that error as I was doing a manual update to prevent the watchtower problems.
I’'ve had an occasional error audit, 6 successful and 8 failed in 12 hours. most audits are successful.
Seems your database were corrupted and part of the historic data is lost. It should not affect your node reputation or something like that, but the stat will be not correct.