Closed madman2021 closed 1 year ago
Your pastebin is not working, but one thing looking at the config is that it is generally not recommended to run the detect stream at such high resolution
Ahh was private sorry made it unlisted.
please do not use debug logs, they are only for specific debugging and in this case makes it more difficult to see what the problem is
Scratch that, ok I'll disable debug logging and re-test.
Well it did it already.
[Frigate](http://mad-app:5000/)
[Frigate](http://mad-app:5000/)
[Cameras](http://mad-app:5000/)
[Birdseye](http://mad-app:5000/birdseye)
[Events](http://mad-app:5000/events)
[Storage](http://mad-app:5000/storage)
[System](http://mad-app:5000/system)
[Config](http://mad-app:5000/config)
[Logs](http://mad-app:5000/logs)
[Documentation](https://docs.frigate.video/)
[GitHub](https://github.com/blakeblackshear/frigate)
Logs
Copy to Clipboard
2023-07-30 04:08:08.267061678 [INFO] Starting Frigate...
2023-07-30 04:08:09.382686392 [2023-07-30 04:08:09] frigate.app INFO : Starting Frigate (0.12.1-367d724)
2023-07-30 04:08:09.426061862 [2023-07-30 04:08:09] peewee_migrate INFO : Starting migrations
2023-07-30 04:08:09.431110899 [2023-07-30 04:08:09] peewee_migrate INFO : There is nothing to migrate
2023-07-30 04:08:09.447319188 [2023-07-30 04:08:09] frigate.app INFO : Output process started: 433
2023-07-30 04:08:09.453953338 [2023-07-30 04:08:09] frigate.app INFO : Camera processor started for Door: 437
2023-07-30 04:08:09.458344003 [2023-07-30 04:08:09] frigate.app INFO : Camera processor started for Right: 439
2023-07-30 04:08:09.462702357 [2023-07-30 04:08:09] frigate.app INFO : Camera processor started for Left: 441
2023-07-30 04:08:09.467202326 [2023-07-30 04:08:09] frigate.app INFO : Capture process started for Door: 442
2023-07-30 04:08:09.472353041 [2023-07-30 04:08:09] frigate.app INFO : Capture process started for Right: 446
2023-07-30 04:08:09.477242909 [2023-07-30 04:08:09] frigate.app INFO : Capture process started for Left: 449
2023-07-30 04:08:09.740789670 [2023-07-30 04:08:09] detector.tensorrt INFO : Starting detection process: 432
2023-07-30 04:08:10.203025765 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] Init CUDA: CPU +188, GPU +0, now: CPU 254, GPU 229 (MiB)
2023-07-30 04:08:10.230570658 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : Loaded engine size: 40 MiB
2023-07-30 04:08:10.231153326 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt WARNING : Using an engine plan file across different models of devices is not recommended and is likely to affect performance or even cause errors.
2023-07-30 04:08:10.579102742 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +192, GPU +76, now: CPU 508, GPU 347 (MiB)
2023-07-30 04:08:10.729139922 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] Init cuDNN: CPU +110, GPU +46, now: CPU 618, GPU 393 (MiB)
2023-07-30 04:08:10.737185583 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +40, now: CPU 0, GPU 40 (MiB)
2023-07-30 04:08:10.737440444 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 578, GPU 385 (MiB)
2023-07-30 04:08:10.737593498 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] Init cuDNN: CPU +0, GPU +8, now: CPU 578, GPU 393 (MiB)
2023-07-30 04:08:10.737699241 [2023-07-30 04:08:10] frigate.detectors.plugins.tensorrt INFO : [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +13, now: CPU 0, GPU 53 (MiB)
2023-07-30 04:10:15.967843859 [2023-07-30 04:10:15] frigate.http ERROR : No recordings found for the requested time range
2023-07-30 04:10:15.999135448 [2023-07-30 04:10:15] frigate.http ERROR : Event does not have recordings: 1690624060.64931-uax5tw
2023-07-30 04:10:35.738813082 [2023-07-30 04:10:35] frigate.http ERROR : No recordings found for the requested time range
2023-07-30 04:10:35.778006581 [2023-07-30 04:10:35] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:10:44.150221757 [2023-07-30 04:10:44] frigate.http ERROR : No recordings found for the requested time range
2023-07-30 04:10:44.185816625 [2023-07-30 04:10:44] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:10:45.893718107 [2023-07-30 04:10:45] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:10:45.907413037 [2023-07-30 04:10:45] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:11:00.288626192 [2023-07-30 04:11:00] frigate.http ERROR : No recordings found for the requested time range
2023-07-30 04:11:00.320497561 [2023-07-30 04:11:00] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:11:31.973425408 [2023-07-30 04:11:31] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:11:31.984169543 [2023-07-30 04:11:31] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
2023-07-30 04:13:13.125705600 [2023-07-30 04:13:13] frigate.http ERROR : No recordings found for the requested time range
2023-07-30 04:13:13.173537193 [2023-07-30 04:13:13] frigate.http ERROR : Event does not have recordings: 1690686498.334268-v9awt0
So the problem seems to be that you have active_objects
enabled for recording but there are no active objects so nothing is being recorded
Okay, so is the "person" in the video not an active object, sorry I thought "active_objects" only recorded video that have an "object" in it vs just random motion
active object does not just mean object, it means an object that is moving. In this case this appears to be a false positive (not actually a person) and something that has not moved
Okay, that unfortunelty doesn't seem to explain why this one has no clip
Infact none of these have clips
What is your docker compose? Do you see any recording files in the recordings file?
root@docker:/var/docker_data/frigate/media/recordings# find . -type f | wc -l
3879
3879 files in media
compose =
version: "3.9"
services:
frigate:
container_name: frigate
privileged: true # this may not be necessary for all setups
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable-tensorrt
shm_size: "256mb" # update for your cameras based on calculation above
volumes:
- /etc/localtime:/etc/localtime:ro
- /var/docker_data/frigate/config.yml:/config/config.yml
- /var/docker_data/frigate/media:/media/frigate
- /var/docker_data/frigate/trt-models:/trt-models
- type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
target: /tmp/cache
tmpfs:
size: 1000000000
ports:
- "5000:5000"
- "8554:8554" # RTSP feeds
- "8555:8555/tcp" # WebRTC over tcp
- "8555:8555/udp" # WebRTC over udp
environment:
FRIGATE_RTSP_PASSWORD: "password"
deploy: # <------------- Add this section
resources:
reservations:
devices:
- driver: nvidia
count: 1 # number of GPUs
capabilities: [gpu]
interesting, for now I would suggest setting the 24/7 recording mode to motion to see if that changes anything for the next event
I have the same issue as @madman2021 There are objects detected correctly (persons and bicycles in my case) and all snapshots are working fine, but all clips are not working (same errors shown as the screenshots above)
It was working in older versions (and recording marked as favorite from long ago are still working) of Frigate. I watched older clips and it was last working at 8/6/23, 8:16:50 AM, the next recording is 8/6/23, 12:55:34 AM and that one is broken (snapshot works, clip broken) and all new recordings after that the clips are also broken.
I'm currently running version 0.12.1-367d724 in docker, with a custom go2rtc version of: 1.6.2
My logging shows things like: [2023-08-08 16:00:05] frigate.record WARNING : Discarding a corrupt recording segment: /tmp/cache/voordeur-20230808155954.mp4 but i dont know if this is related to the problem or just something with the current recording.
When i restart frigate clips and snapshots are working again
It means your recordings are not being saved because they're corrupt. Likely an issue with the camera, config.
Ok but why is everything working again after a frigate restart?
Perhaps the recordings work for some time until the camera starts sending bad data or there is some other issue. There is very limited info here to go off
Strange that if the camera data would be the problem that the detections, snapshots and live view are working fine, but the recordings are broken. I use go2rtc for all streams.
Could this have something to do with the /tmp/cache directory? I have the same setup as @madman2021 with a tmpfs filesystem and the error message is showing the /tmp/cache directory. Now i restarted frigate and the records are working again the mentioned error message seems to be gone as well.
Is there some information i can obtain so it will make troubleshooting easier?
I checked again this morning, recording was working after the restart till 8/8/23, 9:35:47 PM (this recording works) and the next recording at 8/8/23, 11:17:52 PM and all recording after that are broken but the snapshots are fine and showing people.
The Frigate logfiles does not show any errors for the "voordeur" camera at that moment, but when i login to frigate webpage (http://192.168.70.100:5000/) at this moment it shows "No frames have been received, check the error logs" but when i click on the camera (http://192.168.70.100:5000/cameras/voordeur) it is showing the current live view and seems to work fine.
go2rtc logging shows:
"2023-08-08 17:53:37.132063924 17:53:37.131 WRN github.com/AlexxIT/go2rtc/internal/streams/producer.go:171 > error="read tcp 172.18.0.4:33236->192.168.50.116:554: i/o timeout" url=rtsp://admin:PASSWORD@192.168.50.116:554/h264Preview_01_main
2023-08-08 17:53:37.141085046 17:53:37.140 WRN github.com/AlexxIT/go2rtc/internal/streams/producer.go:171 > error="read tcp 172.18.0.4:33220->192.168.50.116:554: i/o timeout" url=rtsp://admin:PASSWORD@192.168.50.116:554/h264Preview_01_sub"
But all detections/clips and snapshots around that time (8/8/23, 5:52:03 PM and 8/8/23, 6:07:14 PM) are working fine.
Later (without reboot in between) it shows this:
"2023-08-08 22:53:11.072743954 22:53:11.072 WRN github.com/AlexxIT/go2rtc/internal/streams/producer.go:171 > error="read tcp 172.18.0.4:34472->192.168.50.116:554: i/o timeout" url=rtsp://admin:PASSWORD@192.168.50.116:554/h264Preview_01_main"
All detections after that time (8/8/23, 11:17:52 PM, 8/8/23, 11:18:07 PM and 8/8/23, 11:57:39 PM) only show snapshots, no clips, and there are no detections/clips/snapshots of today (2023-08-09) but i know for sure there should be a lot of new detections.
Only thing i see is that it are different streams and at the first time it shows 2 rules with both streams (but eveything seems to work as expected) and at the second it is only 1 stream (and after that snapshots work, clips not, but the next day nothing seems to work except live view).
Frigate is using this configuration file and the mentioned camera is called "voordeur":
# MQTT settings
mqtt:
host: 192.168.70.100
port: 1884
topic_prefix: frigate
client_id: frigate
user: frigate
password: "PASSWORD"
stats_interval: 60
# Detectors
detectors:
coral:
type: edgetpu
device: usb
# cpu1:
# type: cpu
# Live view settings
live:
height: 720
quality: 8
ffmpeg:
output_args:
record: preset-record-generic-audio-copy
# hwaccel_args:
# - -hwaccel
# - qsv
# - -qsv_device
# - /dev/dri/renderD128
# Detection stream (motion detection and snapshots)
# NOTE: Can be overridden at the camera level
detect:
width: 1280
height: 720
fps: 5
# Object detection
# NOTE: Can be overridden at the camera level
objects:
track:
- person
# Record configuration
# NOTE: Can be overridden at the camera level
record:
enabled: True
retain:
days: 14
mode: motion
# Event configuration
events:
pre_capture: 5
post_capture: 10
retain:
default: 28
mode: active_objects
# Snapshot configuration
# NOTE: Can be overridden at the camera level
snapshots:
enabled: True
timestamp: False
bounding_box: True
crop: False
retain:
default: 28
go2rtc:
streams:
woonkamer:
- rtsp://CamViewer:{FRIGATE_RTSP_PASSWORD}@192.168.50.110:554/stream1
#- "ffmpeg:woonkamer#audio=opus"
woonkamer_sub:
- rtsp://CamViewer:{FRIGATE_RTSP_PASSWORD}@192.168.50.110:554/stream2
#- "ffmpeg:woonkamer_sub#audio=opus"
keuken:
- rtsp://CamViewer:{FRIGATE_RTSP_PASSWORD}@192.168.50.111:554/stream1
#- "ffmpeg:keuken#audio=opus"
keuken_sub:
- rtsp://CamViewer:{FRIGATE_RTSP_PASSWORD}@192.168.50.111:554/stream2
#- "ffmpeg:keuken_sub#audio=opus"
voordeur:
- rtsp://admin:PASSWORD@192.168.50.116:554/h264Preview_01_main
#- "ffmpeg:voordeur#audio=opus"
voordeur_sub:
- rtsp://admin:PASSWORD@192.168.50.116:554/h264Preview_01_sub
#- "ffmpeg:voordeur_sub#audio=opus"
webrtc:
candidates:
- 192.168.70.100:8555
- stun:8555
# Camera configuration
cameras:
# Woonkamer camera
woonkamer:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/woonkamer_sub
input_args: preset-rtsp-restream
roles:
- detect
- path: rtsp://127.0.0.1:8554/woonkamer
input_args: preset-rtsp-restream
roles:
- record
motion:
mask:
- 465,0,0,0,0,52,465,54
best_image_timeout: 30
mqtt:
enabled: True
timestamp: False
bounding_box: True
crop: False
# Keuken camera
keuken:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/keuken_sub
input_args: preset-rtsp-restream
roles:
- detect
- path: rtsp://127.0.0.1:8554/keuken
input_args: preset-rtsp-restream
roles:
- record
motion:
mask:
- 465,0,0,0,0,52,465,54
best_image_timeout: 30
mqtt:
enabled: True
timestamp: False
bounding_box: True
crop: False
# Voordeur camera
voordeur:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/voordeur_sub
input_args: preset-rtsp-restream
roles:
- detect
- path: rtsp://127.0.0.1:8554/voordeur
input_args: preset-rtsp-restream
roles:
- record
motion:
improve_contrast: True
mask:
- 234,0,0,0,0,42,234,42
best_image_timeout: 30
detect:
width: 640
height: 480
fps: 10
objects:
track:
- person
- bicycle
mqtt:
enabled: True
timestamp: False
bounding_box: True
crop: False
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
I'm with the same problema, but the error apears only in webbrowser in PC (Browser Brave in Linux).... in my android cellphone (inside HA Companion) works fine. when i download this clip, the filename came with no known extension file, so i guess it could be something related to this.
Describe the problem you are having
For whatever reason I seem to get a lot of events that have no clips.
In then makes the event as no clip so hides the clip, this is also confirmed in the logs saying it was not found.
Version
0.12.1-367d724
Frigate config file
Relevant log output
FFprobe output from your camera
Frigate stats
Operating system
Debian
Install method
Docker Compose
Coral version
CPU (no coral)
Network connection
Wired
Camera make and model
reolink,annke
Any other information that may be helpful
No response