| 13 Feb 2025 |
motiejus | I can't get the detect process to read from go2rtc, always weirdly get the same ffmpeg error Error opening input files: Invalid data found when processing input when the detect role comes not directly from camera, but from go2rtc (record is fine).
I will put this on hold for some time (a few days?). I disabled detect process, will be recording only for now (no cpu use).
| 20:37:02 |
@hexa:lossy.network | ERROR : The following ffmpeg logs include the last 100 lines prior to exit.
ERROR : [AVHWFramesContext @ 0x7fd4c0078880] Failed to sync surface 0x9: 1 (operation failed).
ERROR : [hwdownload @ 0x7fd4cc0025c0] Failed to download frame: -5.
ERROR : [vf#1:0 @ 0x1b6ef480] Error while filtering: Input/output error
ERROR : [vf#1:0 @ 0x1b6ef480] Task finished with error code: -5 (Input/output error)
ERROR : [vf#1:0 @ 0x1b6ef480] Terminating thread with return code -5 (Input/output error)
| 20:37:59 |
@hexa:lossy.network | this is what I'm seeing every few minutes | 20:38:03 |
motiejus | also, I can't understand the relationship of go2rtc config in frigate, and its own. I guess I shouldn't use go2rtc top-level in frigate config, and just use the rtsp endpoints. But then the docs don't help understand this too:
After adding this to the config, restart Frigate and try to watch the live stream for a single camera by clicking on it from the dashboard. It should look much clearer and more fluent than the original jsmpeg stream.
| 20:39:05 |
@hexa:lossy.network | frigate can run its own go2rtc process and configure it | 20:39:46 |
motiejus | if I run ffmpeg command myself, I logs lots of lost packets (the router behind the camera is under stress). I think it makes sense to restart in higher-bandwidth situation | 20:39:53 |
@hexa:lossy.network | in nixos I just configure go2rtc myself | 20:39:54 |