Sonarr version (exact version): 4.0.4.1491 Mono version (if Sonarr is not running on Windows): on windows OS: Windows 11 Pro Debug logs: Sonarr Trace Logs Description of issue:
About 1 minute after grabbing a download, tv-sonarr uses 100% of my CPU and doesn’t stop. My computer literally grinds to a halt. Disc response time can be 3-5 seconds. I manually end Sonarr.exe and it is fine until it grabs a download again.
Radarr has the same issue, 100% CPU usage on radarr.exe after grabbing a release.
Both Sonarr and Radarr start the download and import the file just fine. Radarr Trace Log
I have tried deleting files that were 0 bytes in size (corrupted?). I have also tried disabling authentication for localhost. Radarr, Sonarr, qbit are all up to date as well (qbit on v4.6.4). I have also tried stopping bazarr and Ombi in Task Manager to see if that was an issue.
I also use a VPN, Private Internet Access. Sonarr, Radarr, etc are in a split tunnel to use only the VPN, other apps will bypass the VPN. qBit has it’s Network Interface set to the PIA VPN (wgpia0).
All of the tasks under “System → Tasks” run fine as well.
Not sure what else to try, thanks for the help!
Here’s a snippit of the log for those searching for the error message:
|Info|Sonarr.Http.Authentication.BasicAuthenticationHandler|Basic was not authenticated. Failure message: Authorization header missing.
Update - following some Reddit advice, I used a program to find corrupted video files. I used a program called HBBatchBeach. It runs a quick Handbrake encode to see if it passes or throws an error. I scanned ~11,000 files and found ~1000 that were corrupted.
Note for future readers… Normal video files should show a thumbnail in Windows Explorer, corrupted videos do not. You can use HBBatchBeast to find them all, or open folders and look manually.
I will try to run Sonarr/Radarr again and see if the issue persists…
What is your hardware? If you are having corrupted files this points too storage failure and that is why u get high cpu load since it is trying to move bad sectors to another location have you checked the S.M.A.R.T status of your disks ?
Hi, hardware should be solid.
Is there any history or other examples of Sonarr using 100% without an end? I can’t find similar issues online, or figure out why Sonarr/Rardarr are doing this.
AMD 5800X
4080
32GB RAM, DDR4
Plex installed & metadata on M.2 NVME
Media stored entirely on a Seagate Exos 18TB HDD
Picture of SMART status is attached. I don’t see any issues
How does that seem ok to you You have 01- read errors way above the normal and C6 you have a lot of bad sectors that can’t be moved. Your hard is going to die that’s for sure, I suggest you start moving data in small batches to see which sectors are bad also have you done defragmentation on this HDD ever ? If not start with that but yes this HDD is going to die soon. Those big HDDs don’t really like the NTFS system you will be better off making yourself a Truenas machine with some old hardware and at least 2 HDDs for redundency.
I appreciate the advice. Critical files are duplicated and I have a backup with Backblaze as well. Good enough for me- I hope!
Regarding Sonarr, I don’t think this would cause this issue for Sonarr/Radarr, and if so, I wouldn’t be the first to find them in this way, HDDs regularly die.
Regarding SMART, not very knowledgeable here, I did some more reading just now. Seagate Exos seem to have a high Seek Error Rate compared to other vendors. C6 looks okay though, no? Current/worst/Raw = 100/100/0. There are no reallocated sectors either (05). Defrag was gone a few weeks ago. Stablebit Scanner runs in the background as well.
Would love to move to a dedicated server, but just not practical at the moment.
I should add this. Task manager shows 100% usage as “Movies.exe” and/or “TV Shows.exe”. I can’t figure out what these are or what they do in order to troubleshoot. I assume it has —
You know what I might have just figured it out.
I checked the properties → details for the .exes and they showed the original filename as “xmri.exe.” I Googled this and it looks like its’ a crypto mining virus! I’ll try removing this later today… it’s 1:00am now. It must have snuck past Windows Defender & Malwarebytes scans. Probably my fault somewhere along the line.
You have 01- read errors way above the normal and C6 you have a lot of bad sectors that can’t be moved. Your hard is going to die that’s for sure, I suggest you start moving data in small batches to see which sectors are bad also have you done defragmentation on this HDD ever ? If not start with that but yes this HDD is going to die soon. Those big HDDs don’t really like the NTFS system you will be better off making yourself a Truenas machine with some old hardware and at least 2 HDDs for redundency.