[Support]: WARNING : Unable to keep up with recording segments in cache for [camera] #10123
Replies: 1 comment 7 replies
-
|
Make sure you're using the recommended Reolink config as listed in the docs. |
Beta Was this translation helpful? Give feedback.
-
|
Certainly, screenshots are below |
Beta Was this translation helpful? Give feedback.
-
|
you may need to tune motion detection, can see some cameras with skipped fps https://docs.frigate.video/configuration/motion_detection |
Beta Was this translation helpful? Give feedback.
-
|
So motion detection could be the issue? I’ll have to tune it and see then. I know motion detection got a bit more sensitive with 0.13 so maybe it has something to do with that. However the dead camera is indoors so it shouldn’t have to do any heavy lifting in that sense. |
Beta Was this translation helpful? Give feedback.
-
|
I've tuned motion detection so it doesn't register every OSD second change and a few other things in the environment. It now has no skipped frames. But the problem for the camera is persistent. When an other camera gets triggered it writes the file in 0.12 seconds as the following log shows. 2024-03-15 14:06:23.402763774 [2024-03-15 14:06:23] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for Allrummet. Keeping the 6 most recent segments out of 7 and discarding the rest... So the problem lies with that one camera. I've tried to rule out other causes by bypassing switches and changing cables, but the problem is still there. It seems as it is only the sub feed that is not showing as I can live view the main feed from the camera and go2rtc has no error in the logs. That camera works well and can record both feeds from go2rtc in BlueIris so my problem is isolated to my Frigate setup, (BlueIsris records on an other disc). Are there more things I can try? Changing the resolution or bitrate don't improve anything so I guess it's not a network congestion thing? Not that I have a massive crappy network so it should be a problem. |
Beta Was this translation helpful? Give feedback.
-
|
Will be good to get more logs of these recordings, as well as an updated system screenshot |
Beta Was this translation helpful? Give feedback.
-
Describe the problem you are having
I've had Frigate running for quite some time now and it's been working fine until I decided to upgrade from 0.12 to 0.13.1. The first few days it was alright but then one camera ( a Reolink RLC-520) named "Allrummet" in the config.yml started showing no frames received. Now I can't get it back up. The config.yml hasn't changed from 0.12, I already had detect - hight/width configured. I changed the paths in the docker compose file as per instructions for 0.13.
Most likely there's a mistake by myself somewhere that wasn't a problem before the upgrade. But what and where? Upgrading to 0.13.2 didn't make any difference.
I've gone through but cant find anything wrong as the logs just repeating the same message about "Unable to keep up with recording segments in cache"
Version
0.13.2
Frigate config file
Relevant log output
FFprobe output from your camera
Frigate stats
No response
Operating system
HassOS
Install method
Docker Compose
Coral version
USB
Network connection
Wired
Camera make and model
Reolink RCL-520
Any other information that may be helpful
Frigate installed in docker in LCX container in Proxmox. Running Debian 12.5 in the container. Other cameras of same brand and model are on the same VLAN with the same settings in the cameras and firewall and working fine.
Docker compose file:
/Frigate is a separate 1 TB disk mounted in the LXC container. Free disk space right now 850 GB. Free disk space in rootfs about 5 GB out of 20. So no NFS or SMB shares are used.
The LXC's container .conf looks like this:
CPU level is for the Proxmox node 20 % and for the Frigate docker container about 50 % now when this error has come, usually much lower.
Beta Was this translation helpful? Give feedback.
All reactions