Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

Is Nvidia (Frigate?) freezing my system? How to get proper diagnostics?

Featured Replies

My Unraid server is freezing completely, and I believe the issue is related to the Nvidia GPU and the Frigate containers, which use the GPU.
 

System Information:

  • Server Model: Lenovo P910 (recently exchanged due to previous issues)
  • GPU: Nvidia Quadro M2000
  • Unraid Version: 6.12.6
     

Timeline of Events:
 

Hardware Exchange:

  • Recently exchanged Lenovo P910 server due to previous issues.
  • Everything worked fine initially after the exchange.
     

Cache Drive Crash:

  • Cache drive crashed, causing all Docker containers to be disorganized (The mover wasn't properly configured to move Appdata to the array)
     

Reconfiguration:

  • Set up all Docker containers again. Everything worked fine initially.
  • Nvidia driver was auto-updated, which caused the Quadro M2000 not to be recognized.
  • Downgraded the driver, and everything seemed to work fine again.


Server Freezing:

  • After a few days, the server started freezing completely (no more WebUI access).
  • The server ran for about a week without issues when the Frigate container wasn't started.
  • Now in the test starting the Frigate container, the server immediately froze. (Usually the server runs some days, before not being accessible)
  • Sidenote: I need to start the container twice because the USB enumerator changes (this wasn't necessary before the cache drive failure).


Issues:

  • Complete server freeze after starting the Frigate container.
  • Unable to access WebUI when the server freezes.
  • Difficulty obtaining a proper diagnostics file as the device completely freezes, and after a hard shutdown, only logs from the current session are available in the diagnostics file.


Request:
Can you help me troubleshoot this issue?
Any advice or guidance on how to resolve the freezing problem or obtain more useful diagnostic information would be greatly appreciated.


Additional Information:

Attached is a screenshot showing the need to start the Frigate container twice due to USB enumerator changes.


Attached 2 diagnostics - one before the container was started - one after the system hang again and was hard shut down

Image 260.png

tower-diagnostics-20240728-2047.zip tower-diagnostics-20240728-2017.zip

  • Author

just happened again - cannot even SSH into the server - no idea how to get logs in such a state?

  • 4 weeks later...
  • Author

so sad there is absolutely no replies here ...

 

just replaced my USB Stick as it was also being disconnected/lost from time to time ...

 

was hoping this fixes it - today I lost the Unraid USB stick again ... that's so frustrating ...

 

 

The full Lenovo ThinkStation P910 which Unraid is running on has recently also been changed .. can't think it is the machine causing those problems?!

  • Community Expert

Is anything else using your GPU? Have you tried asking in the container support thread/forum?

 

Have you mirrored your syslog to flash so it's saved before it locks?

Edited by Michael_P

  • Author
On 8/25/2024 at 12:48 PM, Michael_P said:

Is anything else using your GPU? Have you tried asking in the container support thread/forum?

 

Have you mirrored your syslog to flash so it's saved before it locks?

No, actually Frigate is the only thing using my GPU - I've also changed it (to the same model which I got here multiple times) ..

 

Is there any guidance to mirror the syslog to flash? This could eventually be really helpful ..

  • Author

it happened again - USB device is lost (see attached screenshot)

/boot is empty as there is no USB device attached any longer

 

after rebooting Unraid I've found this information in syslog - no clue what to do with it

 

alot of mover entries like

 

Quote

Aug 28 03:43:22 Tower root: skip: /mnt/cache/appdata/influxdb2/engine/wal/25399bf07c4ffab4/autogen/32/_00003.wal
Aug 28 03:43:22 Tower root: skip: /mnt/cache/appdata/influxdb2/engine/wal/25399bf07c4ffab4/autogen/33/_00001.wal
Aug 28 03:43:22 Tower root: skip: /mnt/cache/appdata/unifi-network-application/logs/access.log
Aug 28 03:43:22 Tower root: file: /mnt/cache/appdata/unifi-network-application/logs/tasks.log.1
Aug 28 03:43:23 Tower root: skip: /mnt/cache/appdata/unifi-network-application/logs/tasks.log
Aug 28 03:43:23 Tower root: skip: /mnt/cache/appdata/frigate/frigate.db-wal
Aug 28 03:43:23 Tower root: skip: /mnt/cache/appdata/frigate/frigate.db-shm
Aug 28 03:43:23 Tower root: mover: finished

 

some HDD entries about Smart and energy saving

 

Quote

Aug 28 11:55:03 Tower emhttpd: spinning down /dev/sdd
Aug 28 11:55:13 Tower emhttpd: spinning down /dev/sdb
Aug 28 11:56:10 Tower emhttpd: read SMART /dev/sdd
Aug 28 11:56:20 Tower emhttpd: read SMART /dev/sdb
Aug 28 12:26:03 Tower emhttpd: spinning down /dev/sdd
Aug 28 12:26:13 Tower emhttpd: spinning down /dev/sdb
Aug 28 12:27:11 Tower emhttpd: read SMART /dev/sdd
Aug 28 12:27:20 Tower emhttpd: read SMART /dev/sdb

 

 

nothing about any error - probably loglevel is set wrong or the entry is out of scope due to all those mover messages logged?

Image 262.png

Image 263.png

  • Community Expert

The snippets posted should not be relevant, post the complete persistent syslog if you have it.

  • Community Expert

The one that matters is the syslog-previous, assuming that one was after a crash, but unfortunately there's nothing relevant logged, this can be a hardware issue, one thing you can try is to boot the server in safe mode with all docker containers/VMs disabled, let it run as a basic NAS for a few days, if it still crashes it's likely a hardware problem, if it doesn't start turning on the other services one by one.

  • Author
3 hours ago, JorgeB said:

The one that matters is the syslog-previous, assuming that one was after a crash, but unfortunately there's nothing relevant logged, this can be a hardware issue, one thing you can try is to boot the server in safe mode with all docker containers/VMs disabled, let it run as a basic NAS for a few days, if it still crashes it's likely a hardware problem, if it doesn't start turning on the other services one by one.

I can be unlucky of course .. I was using a Lenovo P910 before - had issues (no longer booting - hard freezing) - exchanged the whole thing ...
How are the odds the next Lenovo P910 has hardware issues again?

I had frigate stopped for a while - did not encounter issues then ...


as I got this USB device identitiy (the google coral adapter) switching all the time - could it be, that this also affects the usb stick Unraid is running on?

Eventually some USB setting in the UEFI.

Did you look into the diagnostics? I've searched with visual studio code through all files at once - there are alot of "usb" entries - nothing that shouts ERROR to my face - but as I said before - not sure what I'd be searching for

  • Community Expert
3 hours ago, nOerkH said:

Did you look into the diagnostics?

I did now, and I do see macvlan call traces there, curiously, there weren't any in the syslog-previous, suggesting it's not complete, in any case, macvlan call traces are known to crash servers, update to 6.12.13 since that issue is resolved since 6.12.11, and then retest.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.