WebUI Restore tab shows no files for some clients.

[Dan Broscoi]

Hi, 

On a fresh new server (just cleaned up databases and previously saved backups), the scheduled jobs run, all successful. 

Upon checking the restore tab from UI, some of them present no files (the loading circle goes for a while and then there are no updates on the screen.

Logs show nothing, postgres process CPU stays 100% when the "loading circle" goes and it goes idle.

Example, for a client I got immediately a blank restore window while for other clients it takes a couple of seconds to show nothing.

Where should I look to investigate this ?

[Andy Bird]

sometimes you just need to check back in 5 and the files are present (webui times out but processes continue in the back ground)

You can set a cron to rebuild the catalog each night so that all files are present in the webui when you check

[Dan Broscoi]

vineri, 31 ianuarie 2020, 11:03:53 UTC+2, Andy Bird a scris:

sometimes you just need to check back in 5 and the files are present (webui times out but processes continue in the back ground)

You can set a cron to rebuild the catalog each night so that all files are present in the webui when you check

Well, I checked back in 5, and nothing showed.

I created a new backup:

 

And the restore window still shows nothing after many minutes: 

[Dan Broscoi]

I think I may found the problem, probably it's a bug.

So, like stated before, going to Restore tab -> job name = no results on the box where the file list should be.

If I go on the Job and on the last job of that client I click on the restore icon, it redirects me to the restore tab but this time the files are there.

So:

Restore -> Client -> NO files

Jobs -> backup-job -> restore -> OK files.

Maybe it's something with a previous job because when I select NO to "Merge all client filesets" then I see the files. 

[Andy Bird]

does running

echo ".bvfs_update" | bconsole

make any difference?

[Dan Broscoi]

Did not make any difference.

Anyway, the problem may have some other origins - I have at first a couple of failed jobs due to a typo in the config

Something is weird with this host because the incrementals are 1.3 GB and I know for sure that there's no way to be that much data from a day to day workflow. Maybe a couple of Kb but nothing close to GB.

I do also classic mysqldumps for that instance and the archives are the same, the increase is in size is kilobytes. 

For now i've removed everything and start from scratch to see if the issue persists.