frigate: [Config Support]: v12.x possibly causing HA core to crash

Describe the problem you are having

It appears as though the Frigate add-on crashes and then for some reason, that crashes the deCONZ add-on and then HA core crashes too all the way to the point where I can’t even ping HA any more so the ability to SSH and check or restart anything has gone.

This has happened about 5~6 times over the last 12 months on two different hardware platforms, with the last three events happening in the past 7 days on my new 12th gen Core i5 Beelink SEi12. In typically ‘smoke detector battery warning’ fashion, the last two episodes happened at 0400 and 0200. 🙁 The previous hardware was an Intel NUC with a 5th gen core i7.

The only way to recover HA is to force shutdown the hardware and reboot. Arghh!!!

I have an Uptime Kuma monitor on my Synology NAS that polls the MQTT topic frigate/available and checks for the payload online which is the only way I get notified something has failed. I’ve since added another monitor to ping HA itself.

I have disabled Frigate for now as I have things like my hot water managed via HA and when HA fails and we have no water, I start walking with a limp!

Version

12

Frigate config file

## https://docs.frigate.video/configuration/index#full-configuration-reference ##
## Full configuration File ##
logger:
  # Optional: default log level (default: shown below)
  default: info
  # Optional: module by module log level configuration
  logs:
    frigate.app: debug

mqtt:
  host: 10.0.1.100

detectors:
  coral:
    type: edgetpu
    device: usb

rtmp:
  # Optional: Enable the RTMP stream (default: True)
  enabled: false

birdseye:
  enabled: True
  width: 1280  
  height: 720
  quality: 8
  mode: objects # motion
 
ffmpeg:
  hwaccel_args: preset-vaapi #-hwaccel vaapi -hwaccel_device /dev/dri/renderD128 -hwaccel_output_format yuv420p

detect:
  width: 1280 # <---- update for your camera's resolution
  height: 720 # <---- update for your camera's resolution
  fps: 6

record: # <----- Enable recording
  enabled: True
  events:
    #max_seconds: 300 # default value = 300
    pre_capture: 2 # default = 5
    post_capture: 5
    objects:
      - person
      - cat
    required_zones: []
    retain:
      default: 7 # default = 10

snapshots: # <----- Enable snapshots
  enabled: True
  bounding_box: True 
  # Optional: crop the snapshot (default: shown below)
  crop: False
  # Optional: height to resize the snapshot to (default: original size)
  #height:    

# Optional: in-feed timestamp style configuration
# NOTE: Can be overridden at the camera level
timestamp_style:
  # Optional: Position of the timestamp (default: shown below)
  #           "tl" (top left), "tr" (top right), "bl" (bottom left), "br" (bottom right)
  position: "bl"
  # Optional: Format specifier conform to the Python package "datetime" (default: shown below)
  #           Additional Examples:
  #             german: "%d.%m.%Y %H:%M:%S"
  format: "%d.%m.%Y %H:%M:%S"
  # Optional: Color of font
  color:
    # All Required when color is specified (default: shown below)
    red: 255 # default = 255
    green: 255 # default = 255
    blue: 255 # default = 255
  # Optional: Line thickness of font (default: shown below)
  thickness: 2 # default = 2
  # Optional: Effect of lettering (default: shown below)
  #           "solid" (solid background in inverse color of font)
  #           "shadow" (shadow for font)
  effect: solid

go2rtc:
  streams:
    driveway_frigate:
      - rtsp://username:password@10.0.30.14:554/Streaming/Channels/103 #(3rd stream h.264, 720p, 6fps)
    entry_frigate:
      - rtsp://username:password@10.0.30.12:554/Streaming/Channels/103
    path_frigate:
      - rtsp://username:password@10.0.30.13:554/cam/realmonitor?channel=1&subtype=02 #(3rd stream h.264, 720p, 6fps)
    deck_frigate:
      - rtsp://username:password@10.0.30.11:554/cam/realmonitor?channel=1&subtype=02
    spa_frigate:
      - rtsp://username:password@10.0.30.15:554/Streaming/Channels/103
    backyard_frigate:
      - rtsp://username:password@10.0.30.7:554/Streaming/Channels/103

##################    
#### DRIVEWAY ####
##################  

cameras:
  driveway_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/driveway_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record           
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30    
      mask:
        - 1280,282,1219,234,1144,210,982,159,886,117,819,132,592,720,1280,720
        - 601,0,510,73,0,103,0,0
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
        - cat
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 5000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 250000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.5
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.8 
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

  entry_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/entry_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record             
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30        
      mask:    
        - 0,575,733,325,669,0,0,0
        - 709,361,969,347,947,131,691,141
        - 1280,264,1280,720,1053,720,1001,290              
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
        - cat
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 10000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 225000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.5
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.8
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

  path_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/path_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record             
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30        
      mask:    
        - 1035,0,985,569,617,687,385,291,369,0
        - 1243,70,1246,31,961,29,959,69         
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
        - cat
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 5000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 150000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.5
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.8
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

  deck_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/deck_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record             
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30        
      mask:  
        - 724,0,1280,0,1280,402,834,202,802,172,673,135,536,66,329,78,300,0
        - 1280,720,1280,590,1186,621,1156,720   
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 5000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 150000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.5
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.8
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

  spa_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/spa_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record             
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30        
      mask:
        - 666,94,492,117,330,152,0,254,0,169,0,93,0,0,63,0,158,0,332,0,942,0,897,140
        - 1220,0,1216,150,1068,124,1073,0
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 5000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 150000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.7
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.85
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

  backyard_frigate: # <------ Name the camera    
    ffmpeg:
      inputs:
        - path: rtsp://127.0.0.1:8554/backyard_frigate # <--- the name here must match the name of the camera in restream
          input_args: preset-rtsp-restream
          roles:
            - detect
            - record             
    motion:
      # Increasing this value will make motion detection less sensitive
      # threshold: 30
      # default 25
      # Minimum size in pixels in the resized motion image that counts as motion.
      # Increasing this value will prevent smaller areas of motion from being detected. Decreasing will
      # make motion detection more sensitive to smaller moving objects.
      # As a rule of thumb:
      #  - 15 - high sensitivity
      #  - 30 - medium sensitivity
      #  - 50 - low sensitivity
      # contour_area: 35
      # default: 30        
      mask:
        - 1280,0,1280,190,1198,185,1173,202,1023,193,742,181,521,251,507,476,259,561,0,620,0,0,783,0
        - 46,720,424,720,427,674,45,675
    objects:
      # Optional: list of objects to track from labelmap.txt (default: shown below)
      track:
        - person
        - cat
      filters:
        person:
          # Optional: minimum width*height of the bounding box for the detected object (default: 0)
          min_area: 5000
          # Optional: maximum width*height of the bounding box for the detected object (default: 24000000)
          max_area: 225000
          # Optional: minimum score for the object to initiate tracking (default: shown below)
          min_score: 0.5
          # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: shown below)
          threshold: 0.8
    # Optional: Configuration for the jpg snapshots published via MQTT
    mqtt:
      # Optional: Enable publishing snapshot via mqtt for camera (default: shown below)
      # NOTE: Only applies to publishing image data to MQTT via 'frigate/<camera_name>/<object_name>/snapshot'.
      # All other messages will still be published.
      enabled: True
      # Optional: print a timestamp on the snapshots (default: shown below)
      timestamp: True
      # Optional: draw bounding box on the snapshots (default: shown below)
      bounding_box: True
      # Optional: crop the snapshot (default: shown below)
      crop: false
      # Optional: height to resize the snapshot to (default: shown below)
      height: 540
      # Optional: jpeg encode quality (default: shown below)
      quality: 90
      # Optional: Restrict mqtt messages to objects that entered any of the listed zones (default: no required zones)
      required_zones: []

Relevant log output

Nothing relevant. Here are the last few entries of the HA log before it stopped working.
11/05/2023	14:06:29	Error	homeassistant	user	homeassistant	2023-05-11 14:06:28.876 DEBUG (MainThread) [custom_components.frigate] Finished fetching frigate data in 3.708 seconds (success: True)
11/05/2023	14:06:21	Error	homeassistant	user	homeassistant	2023-05-11 14:06:20.876 DEBUG (MainThread) [custom_components.frigate] Finished fetching frigate data in 3.708 seconds (success: True)
11/05/2023	14:06:13	Error	homeassistant	user	homeassistant	2023-05-11 14:06:12.878 DEBUG (MainThread) [custom_components.frigate] Finished fetching frigate data in 3.710 seconds (success: True)

Frigate stats

No response

Operating system

HassOS

Install method

HassOS Addon

Coral version

USB

Any other information that may be helpful

No response

About this issue

  • Original URL
  • State: closed
  • Created a year ago
  • Comments: 26

Most upvoted comments

You could always try adjusting BIOS settings to see if something improves the situation, but if you are on HA OS then yes that is currently the case

+1
NickM-27 on May 23, 2023
Read more comments on GitHub
← frigate: [Config Support]: Updated to v0.13 - broken
frigate: [Config Support]: WARNING : Unable to keep up with recording segments in cache for <camera-name>. Keeping the 6 most recent segments out of 7 and discarding the rest... →