[Support]: Frigate will not start after updating to HAOS 11.5
Describe the problem you are having
Frigate was fully configured and working including now will not start after upgrading HAOS to 11.5
Version
0.13.01
Frigate config file
ffmpeg:
hwaccel_args: preset-vaapi
mqtt:
host: 172.16.254.25
user: *
password: *
detectors:
coral:
type: edgetpu
device: usb
go2rtc:
streams:
pool_sub:
- "ffmpeg:http://172.16.254.40/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=admin&password=*"
pool_main:
- "ffmpeg:http://172.16.254.40/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=admin&password=*#video=copy#audio=copy#audio=opus"
#- rtsp://admin:*@172.16.254.40:554/h264Preview_01_main
# - "ffmpeg:rtsp_cam_sub#audio=opus"
backdeck_main:
- "ffmpeg:http://172.16.254.41/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=admin&password=*#video=copy#audio=copy#audio=opus"
#- rtsp://admin:*@172.16.254.41:554/h265Preview_01_main
#- "ffmpeg:rtsp_cam_sub#audio=opus"
backdeck_sub:
- "ffmpeg:http://172.16.254.41/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=admin&password=*"
frontyard_sub:
- "ffmpeg:http://172.16.254.42/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=admin&password=*#video=copy#audio=copy#audio=opus"
frontyard_main:
- "ffmpeg:http://172.16.254.42/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=admin&password=*#video=copy#audio=copy#audio=opus"
#- rtsp://admin:*@172.16.254.42:554/h265Preview_01_main
#- "ffmpeg:rtsp_cam_sub#audio=opus"
Doorbell_main:
- "ffmpeg:http://172.16.254.220/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=admin&password=*#video=copy#audio=copy#audio=opus"
Doorbell_sub:
- "ffmpeg:http://172.16.254.220/flv?port=1935&app=bcs&stream=channel0_sub.bcs&user=admin&password=*"
cameras:
doorbell: # <------ Name the camera
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/Doorbell_main?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- record
- path: rtsp://127.0.0.1:8554/Doorbell_sub?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- detect
detect:
fps: 5
enabled: True
record:
enabled: True
retain:
days: 1
mode: motion
events:
retain:
default: 1
snapshots: # <----- Enable snapshots
enabled: True
bounding_box: true
retain:
default: 15
objects:
track:
- person
- dog
- cat
live:
stream_name: Doorbell_main
frontyard:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/frontyard_sub?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- detect
- path: rtsp://127.0.0.1:8554/frontyard_main?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- record
detect:
fps: 5
enabled: True
record:
enabled: True
retain:
days: 1
mode: motion
events:
retain:
default: 1
snapshots: # <----- Enable snapshots
enabled: True
bounding_box: true
retain:
default: 15
objects:
track:
- person
- dog
- cat
- bird
live:
stream_name: frontyard_main
backdeck:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/backdeck_sub?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- detect
- path: rtsp://127.0.0.1:8554/backdeck_main?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- record
detect:
fps: 5
enabled: True
record:
enabled: True
retain:
days: 1
mode: motion
events:
retain:
default: 1
snapshots: # <----- Enable snapshots
enabled: True
bounding_box: true
retain:
default: 15
objects:
track:
- person
- dog
- cat
- bird
live:
stream_name: backdeck_main
pool:
ffmpeg:
inputs:
- path: rtsp://127.0.0.1:8554/pool_main?video=copy&audio=aac
input_args: preset-rtsp-restream
roles:
- record
- path: rtsp://127.0.0.1:8554/pool_sub?video=copy
input_args: preset-rtsp-restream
roles:
- detect
detect:
fps: 5
enabled: True
record:
enabled: True
retain:
days: 1
mode: motion
events:
retain:
default: 1
snapshots: # <----- Enable snapshots
enabled: True
bounding_box: true
retain:
default: 15
objects:
track:
- person
- dog
- cat
- bird
live:
stream_name: pool_main
Relevant log output
2024-02-08 03:48:13.369608562 [INFO] Starting go2rtc healthcheck service...
2024-02-08 03:48:18.373232209 2024/02/08 14:48:18 [error] 148#148: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 127.0.0.1, server: , request: "GET /api/version HTTP/1.1", upstream: "http://127.0.0.1:5001version", host: "127.0.0.1:5000"
2024-02-08 03:48:25.434483170 [2024-02-08 14:48:25] peewee_migrate.logs INFO : Starting migrations
2024-02-08 03:48:25.436839852 [2024-02-08 14:48:25] peewee_migrate.logs INFO : There is nothing to migrate
2024-02-08 03:48:25.440167004 [2024-02-08 14:48:25] frigate.app INFO : Recording process started: 839
2024-02-08 03:48:25.441447132 [2024-02-08 14:48:25] frigate.app INFO : go2rtc process pid: 88
2024-02-08 03:48:25.454912639 [2024-02-08 14:48:25] detector.coral INFO : Starting detection process: 849
2024-02-08 03:48:25.458257674 Process detector:coral:
2024-02-08 03:48:25.458341779 [2024-02-08 14:48:25] frigate.app INFO : Output process started: 851
2024-02-08 03:48:25.458452558 [2024-02-08 14:48:25] frigate.detectors.plugins.edgetpu_tfl INFO : Attempting to load TPU as usb
2024-02-08 03:48:25.458540703 [2024-02-08 14:48:25] frigate.detectors.plugins.edgetpu_tfl ERROR : No EdgeTPU was detected. If you do not have a Coral device yet, you must configure CPU detectors.
2024-02-08 03:48:25.459334506 Traceback (most recent call last):
2024-02-08 03:48:25.459345269 File "/usr/lib/python3/dist-packages/tflite_runtime/interpreter.py", line 160, in load_delegate
2024-02-08 03:48:25.459346619 delegate = Delegate(library, options)
2024-02-08 03:48:25.459351142 File "/usr/lib/python3/dist-packages/tflite_runtime/interpreter.py", line 119, in __init__
2024-02-08 03:48:25.459352259 raise ValueError(capture.message)
2024-02-08 03:48:25.459355854 ValueError
2024-02-08 03:48:25.459368552
2024-02-08 03:48:25.459369649 During handling of the above exception, another exception occurred:
2024-02-08 03:48:25.459370357
2024-02-08 03:48:25.459371274 Traceback (most recent call last):
2024-02-08 03:48:25.459386721 File "/usr/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
2024-02-08 03:48:25.459387647 self.run()
2024-02-08 03:48:25.459388757 File "/usr/lib/python3.9/multiprocessing/process.py", line 108, in run
2024-02-08 03:48:25.459389755 self._target(*self._args, **self._kwargs)
2024-02-08 03:48:25.459397377 File "/opt/frigate/frigate/object_detection.py", line 102, in run_detector
2024-02-08 03:48:25.459398443 object_detector = LocalObjectDetector(detector_config=detector_config)
2024-02-08 03:48:25.459399559 File "/opt/frigate/frigate/object_detection.py", line 53, in __init__
2024-02-08 03:48:25.459400514 self.detect_api = create_detector(detector_config)
2024-02-08 03:48:25.459401751 File "/opt/frigate/frigate/detectors/__init__.py", line 18, in create_detector
2024-02-08 03:48:25.459422261 return api(detector_config)
2024-02-08 03:48:25.459423496 File "/opt/frigate/frigate/detectors/plugins/edgetpu_tfl.py", line 41, in __init__
2024-02-08 03:48:25.459425281 edge_tpu_delegate = load_delegate("libedgetpu.so.1.0", device_config)
2024-02-08 03:48:25.459426457 File "/usr/lib/python3/dist-packages/tflite_runtime/interpreter.py", line 162, in load_delegate
2024-02-08 03:48:25.459451937 raise ValueError('Failed to load delegate from {}\n{}'.format(
2024-02-08 03:48:25.459459890 ValueError: Failed to load delegate from libedgetpu.so.1.0
FFprobe output from your camera
Frigate will not start
Frigate stats
Frigate will not start
Operating system
HassOS
Install method
HassOS Addon
Coral version
USB
Network connection
Wired
Camera make and model
Reolink
Any other information that may be helpful
Updated from HAOS 10.4 to 10.5 Frigate would not start after reboot
The coral is not detected. Do you see the coral in the hardware page in HAOS settings
From my research it does not look like the USB Coral show up in the hardware list for HA, The Coral was working fine before the upgrade and no physical changes. I did see the error about the coral so removed it from the config with the same traceback and error accessing port 5001
Restarting my machine -- not just the virtual machine running HAOS with the Frigate add-on, but the actual machine running Proxmox -- solved this issue for me. I can also confirm that my Coral device (PCIE, not USB) does not appear in the Hardware page in settings.
Restarting my machine -- not just the virtual machine running HAOS with the Frigate add-on, but the actual machine running Proxmox -- solved this issue for me. I can also confirm that my Coral device (PCIE, not USB) does not appear in the Hardware page in settings.
Update: Frigate crashed within 12 hours and would not start back up. Restarting the add-on gave the same No EdgeTPU was detected. If you do not have a Coral device yet, you must configure CPU detectors. error as before. When I rebooted the HAOS VM, I received this failure message in Proxmox:
kvm: ../hw/pci/pci.c:1562: pci_irq_handler: Assertion `0 <= irq_num && irq_num < PCI_NUM_PINS' failed. TASK ERROR: start failed: QEMU exited with code 1
To confirm the issue was caused by the Coral, I turned off hardware passthrough for the PCIE device in Proxmox (which I imagine has the same effect as demating the connector from the VM's perspective), and rebooted the VM again, which succeeded this time. But turning PCIE passthrough back on caused another failed start.
Once again the solution was to reboot the host machine. This fixed my issue for now, but it may not be a permanent solution. Hopefully this info is helpful for others experiencing similar issues, and I'm not just hijacking the thread with irrelevant information!
Yep full reboot of the system seemed to bring it back online once i put the Coral config back in. I wonder what got "stuck" for lack of a better terms
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.