Closed deviant77 closed 1 year ago
Unfortunately these logs don't provide any info other than that recordings were having some sort of issue being saved. Frigate saves recordings in the recordings folder based in UTC time so it is not affected by daylight savings time.
I'm wondering if this may be related to go2rtc, but that's only a guess based on the fact that recordings stopped being created for some reason
ERROR : No new recording segments were created for front in the last 120s. restarting the ffmpeg record process..
Unfortunately these logs don't provide any info other than that recordings were having some sort of issue being saved. Frigate saves recordings in the recordings folder based in UTC time so it is not affected by daylight savings time.
I'm wondering if this may be related to go2rtc, but that's only a guess based on the fact that recordings stopped being created for some reason
ERROR : No new recording segments were created for front in the last 120s. restarting the ffmpeg record process..
I remember the previous issues with DST and the move to UTC, so I was surprised at this occurrence!
About ERROR : No new recording segments were created for front in the last 120s. restarting the ffmpeg record process..
...
Looking through the whole log, this repeats during the whole hour:
2023-10-29 01:38:43.503402136 [2023-10-29 01:38:43] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for garden. Keeping the 5 most recent segments out of 6 and discarding the rest...
2023-10-29 01:38:43.504031402 [2023-10-29 01:38:43] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for pond. Keeping the 5 most recent segments out of 6 and discarding the rest...
2023-10-29 01:38:43.514826237 [2023-10-29 01:38:43] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for crossroad. Keeping the 5 most recent segments out of 6 and discarding the rest...
2023-10-29 01:38:49.400909130 [2023-10-29 01:38:49] watchdog.side ERROR : No new recording segments were created for side in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:49.402645797 [2023-10-29 01:38:49] watchdog.side INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:49.403148443 [2023-10-29 01:38:49] watchdog.side INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:49.471721306 [2023-10-29 01:38:49] watchdog.garden ERROR : No new recording segments were created for garden in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:49.472145826 [2023-10-29 01:38:49] watchdog.garden INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:49.472402558 [2023-10-29 01:38:49] watchdog.garden INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:49.500645196 [2023-10-29 01:38:49] watchdog.front ERROR : No new recording segments were created for front in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:49.500885589 [2023-10-29 01:38:49] watchdog.front INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:49.500891142 [2023-10-29 01:38:49] watchdog.front INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:50.637479380 [2023-10-29 01:38:50] watchdog.crossroad ERROR : No new recording segments were created for crossroad in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:50.637809587 [2023-10-29 01:38:50] watchdog.crossroad INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:50.637813462 [2023-10-29 01:38:50] watchdog.crossroad INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:50.975201366 [2023-10-29 01:38:50] watchdog.pond ERROR : No new recording segments were created for pond in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:50.975209671 [2023-10-29 01:38:50] watchdog.pond INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:50.975212614 [2023-10-29 01:38:50] watchdog.pond INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:51.897284185 [2023-10-29 01:38:51] watchdog.garage ERROR : No new recording segments were created for garage in the last 120s. restarting the ffmpeg record process...
2023-10-29 01:38:51.897391135 [2023-10-29 01:38:51] watchdog.garage INFO : Terminating the existing ffmpeg process...
2023-10-29 01:38:51.897513595 [2023-10-29 01:38:51] watchdog.garage INFO : Waiting for ffmpeg to exit gracefully...
2023-10-29 01:38:53.510391890 [2023-10-29 01:38:53] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for garden. Keeping the 5 most recent segments out of 6 and discarding the rest...
2023-10-29 01:38:53.511396327 [2023-10-29 01:38:53] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for pond. Keeping the 5 most recent segments out of 6 and discarding the rest...
2023-10-29 01:38:53.511686296 [2023-10-29 01:38:53] frigate.record.maintainer WARNING : Unable to keep up with recording segments in cache for crossroad. Keeping the 5 most recent segments out of 6 and discarding the rest...
Frigate says there's too many recordings segments in cache for a particular camera (which would indicate recording segments are being put in to cache?), and then 6 seconds later says that no new recording segments were created for that camera in the last 120 seconds, and then 4 seconds later says there's too many segments again!?
I forgot to add that the same thing occurred on another installation with a very similar configuration (including go2rtc) at the same time.
It seems like it's some sort of error in the logic that's checking to make sure recordings are being written correctly. However, they all use .timestamp()
which shouldn't care about daylight savings so that doesn't quite add up
It's daylight savings time here this morning and I just experienced this same issue. Had to restart Frigate to get recordings started again. Version 0.13.2-6476F8A.
weird, no issues here
Describe the problem you are having
At 02:00 this morning the clock went back to 01:00 for the end of Daylight Saving Time.
For one hour, Frigate's log went crazy (6,512 entries) and all recordings were lost, and the memory and CPU usage starting increasing dramatically.
In the recordings folder, the "01" folder is missing, and the "00" folder was created at 01:00:
When I checked on Frigate at around lunchtime, the memory usage of
frigate.recording_manager
was 44.1% of 8 GB (usually ~1.3%), and the CPU usage was 44.3% (usually ~5.9%). I had to restart Frigate to return to normal memory and CPU usage.Version
0.13.0-cd64399 (beta 3)
Frigate config file
Relevant log output
FFprobe output from your camera
Frigate stats
No response
Operating system
HassOS
Install method
HassOS Addon
Coral version
USB
Network connection
Wired
Camera make and model
Annke C500
Any other information that may be helpful
No response