Update Failed and Now Showing Currently Upgrading


#1

Decided to upgrade via the /admin/upgrade and it showed that docker_manager needed upgrade.

Started the first upgrade and got a memory error so had to reboot the instance and upgrade via terminal instead.

Then discourse (e669fed) needed an upgrade, so did that and it froze with no progress bar activity for more than hour.

Rebooted instance and tried ./launcher rebuild app

Went back to /admin/upgrade and it is showing
New Version Available!
Remote Version: 87cfbfa
Last Updated: 5 hours ago
0 new commits
Currently Upgrading…

Swap shows
Filename Type Size Used Priority
/swapfile file 2044 2040 -1

No progress once again. Anything to look at to make this less painful on 1GB instance?


(Jeff Atwood) #2

Hmm, have not seen problems updating on 1gb instance provided swap is there.

Anything else running on the machine?

Maybe reboot, then upgrade Docker from command line, then upgrade Discourse from command line.


#3

There wasn’t anything else running since it was a fresh reboot after the docker_manager upgrade.

Rebooted again and then went back to cancel the ‘currently upgrading’ status to start over again since this was the second time around with the same message. Status bar went by much quicker and didn’t take long. It seems to be less than 5 minutes so hope it was a full upgrade and everything is there.

Not sure why the browser upgrades didn’t go smoothly so hoping the next one will be better.

Thanks!


(Stephen) #4

Have just experienced this issue too.

Rebooted the host, took a new backup and grabbed a local copy, did ./launcher rebuild app and have just reset the upgrade as /admin/upgrade was still reporting an upgrade in progress even after a rebuilt.

Should rebuild not clear the upgrade flag?