[Support]: Frigate constant crashes in LXC #15995
Replies: 4 comments 5 replies
-
|
I think the main recommendation would be to update to a more recent version of frigate as this will have a newer version of go2rtc. You may prefer to move directly to Frigate 0.15 which has a newer version of ffmpeg that some users have reported improved issues with unstable cameras |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the suggestion Nick. I do not have the go2rtc configuration in my config, so would this upgrade make any difference? Or should I add go2rtc config in my config? |
Beta Was this translation helpful? Give feedback.
-
|
It might still make a difference in other ways. With that said, I would suggest using ffmpeg presets and not manual input args. Also, would not recommend overriding the global args |
Beta Was this translation helpful? Give feedback.
-
|
The upgrade to 0.15 with the newer ffmpeg definitely has a good chance of making a difference |
Beta Was this translation helpful? Give feedback.
-
|
I used this script https://community-scripts.github.io/ProxmoxVE/scripts?id=frigate and installed 0.14, but now ran into a new issue: |
Beta Was this translation helpful? Give feedback.
-
|
Running 0.15 beta which seems to have introduced a warning in the logs - or at least I don't remember having it on 0.13 - around SHM size (I had the default 64GB and adjusted - added this parameter in my stack - to 256MB) and its beens running smooth like a kitten. Will keep on monitoring this, but by now (with all of my cameras enabled), previously Frigate would have crashed already. |
Beta Was this translation helpful? Give feedback.
-
|
@NickM-27 I created a new VM where I installed 0.15 Beta from Portainer (cloned existing VM and updated image target), and adjusted the configuration for the TP-Link cameras to use rtsp: Restarts continue to happen but Frigate seems to be recovering well, but get a log of errors in the logs, some cameras stop streaming and Coral speed is sky high: |
Beta Was this translation helpful? Give feedback.
-
|
The randomly crashing keeps going on. So I needed to make a restart of my Proxmox server fool-proof so that my spouse can start the server again without problems. Unfortunately it is not certain which Bus is assigned to my USB Coral TPU and makes the lxc.mount.entry uncertain wether it is correct or not. Therefor I made a script that figures out the bus and then adjusts the line in the lxc config file: nano /usr/local/bin/update_lxc_config.sh# Path to the LXC config file, please adjust to your frigate container number
lxc_config="/etc/pve/lxc/102.conf"
# Check if the last line in the config is already the correct mount entry
last_line=$(tail -n 1 "$lxc_config")
# Remove the last line and append the new mount entry
if [[ "$last_line" =~ lxc.mount.entry:.* ]]; then
# If the last line contains the mount entry, remove it
sed -i '$d' "$lxc_config"
fi
# Append the new mount entry with the correct bus number
echo "lxc.mount.entry: /dev/bus/usb/00$bus_number dev/bus/usb/00$bus_number none bind,optional,create=dir 0,0" >> "$lxc_config"
echo "Successfully updated the LXC config file with bus number 00$bus_number."then make the sh executable with: chmod +x /usr/local/bin/update_lxc_config.sh
then create a service: nano /etc/systemd/system/update_lxc_config.servicewith the following instructions: [Unit]
Description=Update LXC config for Coral TPU
After=network.target
[Service]
Type=oneshot
ExecStart=/usr/local/bin/update_lxc_config.sh
RemainAfterExit=true
[Install]
WantedBy=multi-user.target
then reload and enable the service with the following commands: systemctl daemon-reload
systemctl enable update_lxc_config.service
systemctl start update_lxc_config.service
check the status after a reboot: systemctl status update_lxc_config.servicethis will do the trick for you for the USB Coral TPU passtrough, even if you are not around and your spouse or kids are restarting the proxmox server. |
Beta Was this translation helpful? Give feedback.
-
|
Appreciate the sharing, but I'm running other services in my Proxmox server that I can't take down everytime Frigate stalls. |
Beta Was this translation helpful? Give feedback.
-
Checklist
Describe the problem you are having
I am running Frigate on a Proxmox LXC container with Coral accelerator, and have been experiencing multiple crashes since initial installation for which I would appreciate help on how to best troubleshoot.
I am constantly monitoring memory and logs, so the main actions I have performed were:
I am not able to capture the logs before the error (or at least I don't know how)
Steps to reproduce
...
Version
0.13.2-6476f8a
In which browser(s) are you experiencing the issue with?
Google chrome 131.0.6778.140
Frigate config file
docker-compose file or Docker CLI command
Relevant Frigate log output
Relevant go2rtc log output
Operating system
Proxmox
Install method
Docker Compose
Network connection
Mixed
Camera make and model
Reolink RLC-821A, TP-Link C200, TP-Link C220, TP-Link C520WS
Screenshots of the Frigate UI's System metrics pages
Any other information that may be helpful
This is the LXC summary prior restarting it again:
This is after I reboot the container:
Beta Was this translation helpful? Give feedback.
All reactions