I keep getting the following error several hours in my PMM run: Plex Media Server is currently running database migration
The strange thing is that I'm not doing any database migration. Any idea what may be causing this and how I can fix it?
#ERROR - "Plex Media Server is currently running database migration"
bright sorrel
flint pineBOT
Welcome Back!
Thanks for being a PMM Sponsor, we greatly appreciate it! Your ticket will now be diverted to <@&1097919568334311495>.
Please provide any logs if you think they're relevant to your post.
You can press the "Close Post" button above or type !close at any time to close this post.
bright sorrel
Not sure if it would help but I tried setting timeout: 86400 in my config.yml but that didn't seem to do anything. (Basically tried setting it to a very large value in hopes if there would be an error PMM would attempt to retry)
rare jacinth
Not sure what would be causing that but it's coming from Plex. PMM's just doing a request to get info and Plex is responding with:
<?xml version="1.0" encoding="utf-8"?><Response code="503" title="Maintenance" status="Plex Media Server is currently running database migration."/>
The timeout value from the config only comes into play when the error is actually a timeout. If the server returns an actual error code PMM doesn't try again.
bright sorrel
Hmmm
Makes sense
What do you guys think about the feature request to retry on error as well and not only timeouts?
The idea being if plex goes unhealthy and restarts then PMM can resume where it left off instead of needing to start from the beginning all over again.
rare jacinth
Feels risky to start ignoring errors, but at a minimum this would need a limit on retries.
bright sorrel
Yes, a limit with cool-down period would be required for sure.
rare jacinth
Seems like a bunch of scaffolding for something that probably isn't recoverable in most cases [like an auth error, no point in retrying that], unless the config also requires specifying the errors to retry, which seems problematic now that I think of it since I imagine a lot of different things fall under various error codes.
bright sorrel
Perhaps maybe only target this one since it seems most common 🙃 . I've tried to run PMM 3 times and keep getting this error after 2 days into the run. Each time I hope it's a 1-off but it appears not.
But worst case even if we were to retry all errors with a cooldown period of say 5 minutes between retries and max retries set to something defined in the config, what would be so bad about that? Worst case it doesn't work and fails anyways. (no worse than the current result)
rare jacinth
the "worse" part would be that it would take at least 5 minutes to fail every time this happens, maybe 10 or 15 or longer depending on the retries, so now your 2-day run takes 3 days since it's waiting around for 20 minutes on every error.
Probably the better approach would be to reduce the amount of stuff happening in any given run so it doesn't take 2 days.
bright sorrel
Judging from the number or errors in the logs I don't think it would take an extra day, maybe a few hours though (potentially less depending what value is set for retries and the cooldown period). I'm still of the mindset it would be beneficial since it would save much more time compared to re-running it again from scratch.
Nothing else is running or interrogating plex. Just PMM, autoscan, and some people watching.
And Tautulli.
I guess I can try disabling overseerr.
rare jacinth
Sure, but if the pain point is that you have to start a 2-day run over from scratch, it seems like the simpler approach is "don't do 2-day runs".
LIke that asian movie library of ours doesn't change enough that the country collections would have to be run very often at all.
And not all in a single run..
bright sorrel
After the first run I plan on reverting back to my schedules for each collection. That way runs are only a couple hours in duration for the most part. But when you have 40+ collections across multiple libraries it becomes a pain to edit all the yaml files corresponding schedules on multiple ocassions so I was trying to take a shortcut 😅 . I guess I may need to do that though if all else fails.
rare jacinth
Judging from the number or errors in the logs I don't think it would take an ext...
The issue is that this is a sample size of one. I've seen quite a few logs with quite a few more errors than that, and if they all take an extra 15 minutes it can add up Maybe not an extra day, but..
bright sorrel
Fair point. Perhaps not to enable it by default but provide the option within the config.
rare jacinth
becomes a pain to edit all the yaml files corresponding schedules
It'd be simpler [IMO] to use runtime flags or multiple config files to control this rather than editing schedules inside the YAML.
quasi valve
I have seen the Plex db migration message upon restart of Plex. Usually accompanied by some kind of upgrade…. And essentially Plex is unusable
bright sorrel
Yeah, I think at some point plex is going unhealthy and autoheal restarts it and then PMM chokes. Hence why I think an option retry option in the config would be beneficial.
quasi valve
Since I run PMM daily, thats the retry for me... 😉
bright sorrel
@peak sable Thought on the feature request for the ability to optionally configure retries in the config? 🙃