Radarr - Radarr problems. WebUI unresponsive for minutes at a time. | PlexGuide.com

Radarr Radarr problems. WebUI unresponsive for minutes at a time.

  • Stop using Chrome! Download the Brave Browser via >>> [Brave.com]
    It's a forked version of Chrome with native ad-blockers and Google's spyware stripped out! Download for Mac, Windows, Android, and Linux!
Welcome to the PlexGuide.com
Serving the Community since 2016!
Register Now

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
So.. here's the story.

I've been working on and off on using this project the last 3+ weeks. I recently bought $1200 worth of hardware (Synology NAS and drives) to use Plex Locally. came across plexguide and loved the idea of unlimited storage, sharing with friends/family etc.

Soo..
Now i'm on at least server rebuild #5. I've had hetzner dedicated servers (4) and now a RS 4000 (SAS) from netcup.
the last 3 times i've built from scratch - and always run into problems with radarr. and only radarr it seems.

here's a snippet from my trace level log.


19-1-17 22:23:16.4|Debug|ImportDecisionMaker|Analyzing 0/1 files.
19-1-17 22:23:16.4|Trace|DiskScanService|Import decisions complete for: [It's a Wonderful Life (1946)][tt0038650, 1585] [00:00:00.0039486]
19-1-17 22:23:16.4|Debug|ImportApprovedMovie|Decisions: 0
19-1-17 22:23:16.4|Info|DiskScanService|Completed scanning disk for It's a Wonderful Life
19-1-17 22:23:16.4|Trace|EventAggregator|Publishing MovieScannedEvent
19-1-17 22:23:16.4|Trace|EventAggregator|MovieScannedEvent -> MovieScannedHandler
19-1-17 22:23:16.4|Trace|EventAggregator|MovieScannedEvent -> UpdateMovieFileService

19-1-17 22:30:32.4|Trace|Http|Req: 36 [GET] /api/profile
19-1-17 22:30:32.4|Trace|Http|Req: 37 [GET] /api/rootfolder
19-1-17 22:30:32.4|Trace|Http|Req: 38 [GET] /api/movies/discover/recommendations
19-1-17 22:30:32.4|Trace|Http|Req: 39 [GET] /
19-1-17 22:30:32.4|Trace|Http|Req: 40 [GET] /api/calendar?unmonitored=false
19-1-17 22:30:32.5|Trace|Scheduler|Pending Tasks: 4
19-1-17 22:30:32.5|Trace|Http|Res: 39 [GET] /: 200.OK (65 ms)
19-1-17 22:30:32.5|Trace|CommandQueueManager|Publishing CheckForFinishedDownload
19-1-17 22:30:32.5|Trace|CommandQueueManager|Checking if command is queued or started: CheckForFinishedDownload
19-1-17 22:30:32.5|Trace|CommandQueueManager|Inserting new command: CheckForFinishedDownload
19-1-17 22:30:32.5|Debug|UpdateMovieFileService|Date of file [/mnt/unionfs/movies/It's.a.Wonderful.Life.(1946)/It's.a.Wonderful.Life.(1946).Unknown-FraMeSToR.mkv] changed from '1/1/1970 12:00:00 AM' to '12/20/1946 12:00:00 AM'
19-1-17 22:30:32.5|Debug|UpdateMovieFileService|Changed file date for 1 files of 1 in It's a Wonderful Life
19-1-17 22:30:32.5|Trace|EventAggregator|Publishing CommandUpdatedEvent
19-1-17 22:30:32.5|Trace|Http|Req: 41 [GET] /api/calendar?start=2018-12-30T06%3A00%3A00.000Z&end=2019-02-10T06%3A00%3A00.000Z&unmonitored=false


note the -seven-minute gap in the middle there?
This build of radarr has 43 movies in its collection. only about 20 or so are actually downloaded. This problem happens anytime radarr re-scans it's database or you do an 'update library'.

it will -very- slowly eventually finish walking through all the movies.

If you restart radarr it remembers what it was doing and shortly after the restart gets into this non-responsive state again.

While this is happening, CPU is fine (netdata shows like 50%).
Other processes like PLEX, sonarr, sabnzbd, portainer, netdata are all responding fine.

I don't even know where to begin to try and troubleshoot this. I've turned on trace logging, ffs, and it doesn't give me any details as to why it 'hung' for 7 minutes.

I've got 'analyze video files' set to off. I also tried in a previous attempt at this to try and not have it change the file dates of the movies, but that didn't seem to make a difference.

Always keep running into this 'pause for several minutes at a time' problem.

currently running traefik with portguard and pgshield, but also ran into this before those were turned on. (wasn't able to even use traefik until last week when someone posted the cloudflare workaround -if you bought the domain when you signed up with gsuite)



I'm usually hanging around on discord.

help.
 

plex_noob

Elite
Staff
Donor
Oct 1, 2018
224
113
Have you tried installing netdata container to check the hardware?
Are there any other processes running at the same time?
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
Have you tried installing netdata container to check the hardware?
Are there any other processes running at the same time?
While this is happening, CPU is fine (netdata shows like 50%).
Other processes like PLEX, sonarr, sabnzbd, portainer, netdata are all responding fine.
 

plex_noob

Elite
Staff
Donor
Oct 1, 2018
224
113
This could not only be a CPU problem. What about your disk?
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
disk seems fine - all other processes (radarr, plex, plexguide, docker) seems fine.
also - has happened on 3 completely separate servers, all starting over from scratch with radarr.

Today it behaved very well for the most part.
until I told it to refresh -1- movie. then it hung.
restarted it with docker
finished that refresh - told it to refresh one other movie.
then it hung
then docker couldn't even restart/stop/kill it.
so i rebooted the server...

afraid to tell it to refresh any movies ever again.. (but it will automatically soon i'm sure)
 

doperyde

Experienced
Staff
Mar 7, 2018
53
24
Have you reached out to the radarr devs or searched in their resources?
 
  • Like
Reactions: 1 user

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
i did talk to some people on the radarr discord, but at the time it was hard to dig through all the trace level logs as there was too much going on. I was unable to give them this 'seven minute gap' issue.

I did find a few similar threads on their reddit subforum, but most of them never had an answer. The only ones that 'solved' it were by turning off 'analyze video files', and i've already got that option disabled.

I just don't understand why I'm the only one reporting this.
 

hooper

Governer
Staff
Donor
Aug 1, 2018
310
119
disk seems fine - all other processes (radarr, plex, plexguide, docker) seems fine.
How are you determining that disk is fine? How much RAM does your server have? Are you running the standard PG configuration for Radarr or have you altered the setup in some way?
 

Admin9705

Administrator
Project Manager
Donor
Jan 17, 2018
5,156
2,112
if your radarr is large and huge, it's part of the reason in my experience :D plus if your disk is taxed, it's the biggest problem for it!
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
see above where i say i'm having this problem with only 40 movies in radarr's database.
i'm assuming when you ask about disk you're suspecting bad hardware? Again - 3 seperate servers, 2 different providers (hetzner and now netcup)
netdata doesn't show disk as pegged to the max whenever it does this. also - if it were disk activity, i would expect the other processes to stall as well.
running standard PG configuration. (again, rebuilt server from scratch at least twice)

next time it happens, I'll use docker and shut down all other apps (sonarr, sabnzbd, plex, etc) to see if it matters.
 

Admin9705

Administrator
Project Manager
Donor
Jan 17, 2018
5,156
2,112
see above where i say i'm having this problem with only 40 movies in radarr's database.
i'm assuming when you ask about disk you're suspecting bad hardware? Again - 3 seperate servers, 2 different providers (hetzner and now netcup)
netdata doesn't show disk as pegged to the max whenever it does this. also - if it were disk activity, i would expect the other processes to stall as well.
running standard PG configuration. (again, rebuilt server from scratch at least twice)

next time it happens, I'll use docker and shut down all other apps (sonarr, sabnzbd, plex, etc) to see if it matters.
oh ya? ya mine doesn't stall at all and it's much much bigger. might be your box :D
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
oh ya? ya mine doesn't stall at all and it's much much bigger. might be your box :D
but if it was hardware / the box.. why would the problem persist across 3 different machines. 2 were bare metal/dedicated.. the netcup is an rs4000 g8, which has 32gb ram, 10 dedicated cores, 960gb SAS.. should definitely be beefy enough to handle this.
 

doperyde

Experienced
Staff
Mar 7, 2018
53
24
hmm, have you tried a different web browser? disable any kind of script blocking or ad blocking type of addons if you may be running any? Perhaps try an incognito/in private browsing window?
If the problem seems to persist over multiple servers, etc, maybe it isn't something on the remote side, but the local client side?
just a thought?
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
hmm, have you tried a different web browser? disable any kind of script blocking or ad blocking type of addons if you may be running any? Perhaps try an incognito/in private browsing window?
If the problem seems to persist over multiple servers, etc, maybe it isn't something on the remote side, but the local client side?
just a thought?
Not a bad thought - but I've tried chrome and IE from my machine, and also my work machine.
I've also used chromium from the server itself (using x2go to get a remote GUI to the server).
When it's hung, radarr (and only radarr) is unresponsive. The log files also freeze for these several minutes, and when they start back up they don't give any indication as to what happened, or even that there was a problem (doesn't spit out an error /trace stack, etc).

When it works, it works great. It's just that when it gets stuck, it stops processing things in sab's download folder and that fills up the server drive quick. Until it is done scanning that folder (or all of them) it doesn't process any movies or handles them very slowly.
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
It has hung again this morning.. there are some interesting sqlite errors in the trace log right before it did this.

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|EventAggregator|Publishing MovieScannedEvent
19-1-19 07:52:32.3|Trace|EventAggregator|MovieScannedEvent -> MovieScannedHandler
19-1-19 07:52:32.3|Info|MovieScannedHandler|[The House with a Clock in Its Walls] was recently added, performing post-add actions
19-1-19 07:52:32.3|Trace|CommandQueueManager|Publishing MoviesSearch
19-1-19 07:52:32.3|Trace|CommandQueueManager|Checking if command is queued or started: MoviesSearch
19-1-19 07:52:32.3|Trace|CommandQueueManager|Inserting new command: MoviesSearch
19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|CommandExecutor|MoviesSearchCommand -> MovieSearchService
19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|CommandQueueManager|Marking command as started: MoviesSearch
19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|EventAggregator|MovieScannedEvent <- MovieScannedHandler
19-1-19 07:52:32.3|Trace|EventAggregator|MovieScannedEvent -> UpdateMovieFileService
19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|ConfigService|Using default config value for 'parsingleniency' defaultValue:'Strict'
19-1-19 07:52:32.3|Trace|ConfigService|Using default config value for 'parsingleniency' defaultValue:'Strict'
19-1-19 07:52:32.3|Debug|Parser|Parsing string 'Saw.3.2006.Unrated.DC.1080p.BluRay.DTS.x264-CyTSuNee'
19-1-19 07:52:32.3|Trace|Parser|^(?<title>(?![(\[]).+?)?(?:(?:[-_\W](?<![)\[!]))*(?<year>(19|20)\d{2}(?!p|i|(19|20)\d{2}|\]|\W(19|20)\d{2})))+(\W+|_|$)(?!\\)
19-1-19 07:52:32.3|Debug|Parser|Movie Parsed. Saw 3 - 2006
19-1-19 07:52:32.3|Debug|ParsingService|Language(s) parsed:
19-1-19 07:52:32.3|Debug|QualityParser|Trying to parse quality for Saw.3.2006.Unrated.DC.1080p.BluRay.DTS.x264-CyTSuNee
19-1-19 07:52:32.3|Debug|ParsingService|Quality parsed: Bluray-1080p v1 (Extended or Unrated)
19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on AlternativeTitles(MovieId)

19-1-19 07:52:32.3|Trace|Owin|SQLite warning (284): automatic index on MovieFiles(MovieId)



I've tried shutting down all the other containers I could to see if that made a difference, but it doesn't seem to.

netdata shows disk read/write at 0 mb/s. cpu at 20%, and inbound data is actually hovering between 200-800 megabits/s (which is kind of odd considering i stopped sabnzbd)
 

doperyde

Experienced
Staff
Mar 7, 2018
53
24
I was really bored last night and again tonight and dug around a bit.
Found a couple potential leads regarding/suggesting some strange behaviors potentially being related to kernel bugs..
https://forum.rclone.org/t/rclone-plex-library-scan-extremely-slow-24h/7328/5
https://forum.rclone.org/t/sudden-drop-in-download-bandwidth-during-transfer/7255

A kernel bug could explain why it's followed across multiple servers.
uname -r should show you what kernel you're running on

Another lead is trying to use the browsers developer console to maybe see whats going on
also brings up the fact that the darr apps use mono (open source .net) and are memory hungry.
https://www.reddit.com/r/radarr/comments/6u9ziv
You've mentioned checking system resources, but not mentioned what memory usage looks like. Not doing any kind of memory limiting on anything by chance (or even by accident)?

Instead of netdata, have you checked out cadvisor? it's a gui for docker monitoring and can be installed from pg community box. maybe it will help take a look?



As for the sqlite messages, those aren't 'errors' they are 'warnings' :) I wouldn't worry about them too much.

https://www.sqlite.org/rescode.html
(284) SQLITE_WARNING_AUTOINDEXThe SQLITE_WARNING_AUTOINDEX result code is passed to the callback of sqlite3_log() whenever automatic indexing is used. This can serve as a warning to application designers that the database might benefit from additional indexes.

Basically due to the amount of data the db is dealing with or there not being an index for that data, the db (sqlite) will decide to make an index automatically as its more efficient.
 

DrgnFyre

Experienced
Original poster
Dec 25, 2018
65
8
I was really bored last night and again tonight and dug around a bit.
Found a couple potential leads regarding/suggesting some strange behaviors potentially being related to kernel bugs..
Wow. Good finds.

I had basically given up at this point and actually cancelled my netcup server. Decided to just stick with my local storage.

I was using Ubuntu 18.01 LTS, but they reference that version specifically in your threads.

Now I'm thinking of firing up another Hetzner dedicated to try again.. (man i'm a glutton for punishment) (too bad they don't setup new servers on the weekend)

My deadline is the end of the month - that's how long I have to return my local storage solution. i have no need for both local and remote storage. Remote would give me the ability to share my library with my friends, and i love the idea of unlimited storage.
 

Recommend NewsGroups

      Up To a 58% Discount!

Trending