@OnFreund commented on GitHub (Oct 10, 2023):
I reserver the bad ram reported by memtest to prevent it from being used, yet the problem is still hapenning. A system with bad ram is obviously…
@NickM-27 commented on GitHub (Oct 7, 2023):
Seems like a good place to start would be running a memtest
@OnFreund commented on GitHub (Sep 28, 2023):
Yeah, what I meant is that maybe it should be watched like them. It's an important process.
@NickM-27 commented on GitHub (Sep 28, 2023):
perhaps, the difference is that the ffmpeg processes have an expectation that a crash may occur because of a camera restarting or network timeout or…
@OnFreund commented on GitHub (Sep 28, 2023):
Fair enough. The implications are serious though - you lose all the recordings until you notice the crash.
@OnFreund commented on GitHub (Sep 28, 2023):
Also, don't know if it's related, but now I can't see clips in HA:
Traceback (most recent call last): File "/usr/src/homeassistant/homeassistant…
@OnFreund commented on GitHub (Sep 29, 2023):
Ok, memtest does show a lot of errors - good catch! I'll try to replace the memory modules and hope it gets resolved.
Closing this issue for now,…
@MacieGx commented on GitHub (Oct 7, 2023):
Today I experienced a similar error, I also lost the recordings until I restarted. 😢 The problem occurred on one of the three cameras.
`2023-10-07…
@NickM-27 commented on GitHub (Sep 28, 2023):
I don't think anything is restarting. Based on your description, most likely the seg fault is occurring in the recording maintainer which was part of…
@OnFreund commented on GitHub (Sep 28, 2023):
Got it. So maybe the recording process could be dogwatched? I think that ffmpeg processes are being restarted by Frigate, aren't they?
@NickM-27 commented on GitHub (Sep 28, 2023):
the ffmpeg processes are watched and automatically restarted, the recordings maintainer is not
@OnFreund commented on GitHub (Sep 28, 2023):
Seg fault is back:
frigate
@NickM-27 commented on GitHub (Sep 28, 2023):
Unfortunately segfaults are a somewhat generic error, it is difficult to know what the cause is especially when you are the only user to report this.…
@OnFreund commented on GitHub (Sep 28, 2023):
I'll check that. In the meantime, there does seem to be a regression in 0.13 in how seg faults are treated. In 0.12, docker compose would restart…
@OnFreund commented on GitHub (Sep 27, 2023):
What can cause this and what are the options to repair?
@NickM-27 commented on GitHub (Sep 27, 2023):
sqlite is generally robust with lots of safeguards so I'm not sure what cases would cause this problem, or how to repair besides restoring from a…
@NickM-27 commented on GitHub (Sep 23, 2023):
yes, restreaming (especially high res streams) does use CPU resources
@OnFreund commented on GitHub (Sep 24, 2023):
Some updates:
- There are no crashes in the last 14-15 hours, but a cases of the detection process being stuck and having to be forced quit…
@OnFreund commented on GitHub (Sep 27, 2023):
No crashed for the last 4 days, but now I'm getting these:
2023-09-27 08:26:48.094735187 File "/usr/lib/python3.9/threading.py", line 954, in…
@NickM-27 commented on GitHub (Sep 27, 2023):
looks like the database is corrupt