[Detector Support]: Frigate Crashes Requiring Power Cycle of Coral TPU then LXC Reboot #16988
Replies: 2 comments
-
|
#5392 |
Beta Was this translation helpful? Give feedback.
-
|
Back around the time of my last post I found a thread that suggested adding #backchannel=0 to all the camera feeds. That seemed to actually fix this until a few weeks ago when the issue reared its ugly head again. I just upgraded to Frigate .16 beta 3 to see if it would fix it. It seems to have helped but its still happening. By helping I mean this time when it happened it didn't seem to bring down all the cameras and only affected the cameras intermittently but still required a restart to get the system back to normal. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Describe the problem you are having
Every 12-48hrs(more recently it seems like only 12 hrs) the whole Frigate Stack locks up. If I just reboot the Docker LXC the issue persists. I have to power cycle the Coral TPU by unplugging it from the USB and then plugging it back in and rebooting the Docker LXC to get the system running properly again.
Version
0.15.0-cea210d
Frigate config file
docker-compose file or Docker CLI command
Relevant Frigate log output
Install method
Docker Compose
Object Detector
Coral
Screenshots of the Frigate UI's System metrics pages
Any other information that may be helpful
I thought this may be an issue with the Coral TPU. I had the Coral on a powered USB hub and I've tried connecting it directly to the host machine which is a Lenovo Thinkcentre M910q but that had no affect. I've also tried swapping the USB cable which also had no affect. Regarding the attached error about inprocessed recordings in the cache I've found some mention of a possible network share issue but I'm not sure how to validate if that is the issue or not and I'm not sure why that would require the Coral to be power cycled to fix. The issue does not seem to be tied to happening under heavy detection load or anything like that.
Beta Was this translation helpful? Give feedback.
All reactions