-
Notifications
You must be signed in to change notification settings - Fork 132
-
Notifications
You must be signed in to change notification settings - Fork 132
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Borg subprocess is not started #187
Comments
do you use a ssh: repo? |
No, the repo is on a mounted CIFS share. The repo is working fine (since several months) using a bash script running from a cron job. |
hmm, the log looks like vorta is trying to create new instances (of the scheduler?) and can't. |
I think that's because it thinks that the previous run is still active. The progressbar was still running when I came back to my computer. |
Ok, today it happened again.
I found that the main vorta-process had a subprocess running. After ending that with a SIGQUIT the "list" process was listed with returncode "-3" in the protocol and subsequent backups could run again. I'm not really fluent enough in python to know where to start looking for something like this. Last I asked there was no way to enable something like a debugging log, right? |
I see. Will do some testing on this too. Maybe adding some timeout could fix it. I've never seen a Borg process get stuck, except for an input prompt. |
Well, there was no Borg process. Only vorta. So Borg itself must have terminated at some point but vorta didn't finish whatever it needed to do after that. |
OK. Then the termination isn't seen by |
Must happen in the BorgThread class. Any particular command where it happens? |
No, right now it's the prune command (yes, it's happening again). Mostly it's the list-command. But also during "check". As far as I can see (in the log), it never happened during "create". |
Ok, not sure this will help you in any way but according to gdb it's spinning somewhere inside /usr/lib/python3.6/subprocess.py I've attached to the running vorta sub-process and I get this. As I said, I currently have extremely basic knowledge of Python so I'll refrain from trying to interpret what might be happening here.
|
…g processes when finding mount points. Fixes borgbase#187
I couldn't reproduce this on macOS or Ubuntu (VM), but your stacktrace was very useful nonetheless. The culprit may be the
There seems to be a newer technique to group processes and terminate them at once (like we need when cancelling a running Borg process). I changed it in this PR: #194 Can you try if this fixes the issue for you? The commands to pull my PR would be:
Then (as you probably know already)
|
I didn't know, so thank's for telling me (why did I write asking here? scratchhead). I'm running that now. Let's see what happens in the next hours/days... |
Ok, this modified version has been running for 48 hours (hourly backups) without issue, before the error was intermittent but mostly it occured at least once every 24 hours. So I'll say already, that this is the most stable Vorta version I've had running so far. I'll wait a few more days before marking this as "fixed". On the other hand: Your change definitely didn't introduce other issues here, so it's definitely an improvement. |
Glad to hear. Let's wait a few more days and merge it then. |
If you still didn't encounter a "hang", @JanSchnacki , I'll go ahead and merge it together with 2 other hotfixes we had this week. |
Yes, @m3nu , go ahead. Everything seems to run as smooth as I can wish for, currently. |
Describe the bug
Vorta hangs after some time. Or rather: The GUI still does "something" (responds to clicks up to a point) but backups won't run anymore.
To Reproduce
I have no idea how to reproduce this. It's happening randomly. The log does not show anything that would lead to a possible cause
Desktop (please complete the following information):
Kubuntu 18.10
Additional context
Yesterday evening vorta started a backup at 23:00 which finished around a minute later. The "prune" run then never terminated. But I could not find a running process either. This was the log when I came back to my computer tonight:
The text was updated successfully, but these errors were encountered: