[Detector Support]: Tensorrt version won't start - stuck on starting go2rtc healthcheck service #15288

yangrusen · 2024-12-01T15:04:02Z

yangrusen
Dec 1, 2024

Describe the problem you are having

I encountered a similar issue with #9619, where my Frigate installation is trying to access the internet to download TensorRT models, but my network restrictions prevent this. Since I am in China, Frigate container cannot access the necessary URLs.

Is there a way to download the TensorRT models separately and then upload them to the appropriate folder manually? Any guidance on how to do this would be greatly appreciated!

Version

latest stable-tensorrt

Frigate config file

mqtt:
  host: 192.168.2.2 #mqtt的ip
  port: 1883 #mqtt端口
  user: elvis #mqtt用户名,如果没有的话,可不写
  password: jlksp100 #mqtt密码,如果没有的话,可不写
  client_id: frigate #用于区分

detectors:
  tensorrt:
    type: tensorrt
    device: 0

model:
  path: /config/model_cache/tensorrt/yolov7-320.trt
  input_tensor: nchw
  input_pixel_format: rgb
  width: 320
  height: 320

objects:
  filters:
    person: #这里是设置需要检测的对象,官网有很多种类型,这里以检测人为准
      min_score: 0.7 #这是最小分值,当检测到的物体,超过70%的概率为人体,才确认是人体,这个参数根据自己环境的物体和光线自己调整

cameras:
  camera_1: # <--- this will be changed to your actual camera later
    enabled: true
    record:
      enabled: false #是否开启录制功能
    ffmpeg:
      hwaccel_args: preset-nvidia-h264
      inputs:
        - path: rtsp://admin:[email protected]:554/cam/realmonitor?channel=1&subtype=0
          roles:
            - detect
    detect:
      width: 1280 # <---- update for your camera's resolution
      height: 720 # <---- update for your camera's resolution
      stationary:
        interval: 50
        threshold: 50

docker-compose file or Docker CLI command

docker run
  -d
  --name='frigate2'
  --net='bridge'
  --pids-limit 2048
  --privileged=true
  -e TZ="Asia/Shanghai"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Unraid"
  -e HOST_CONTAINERNAME="frigate2"
  -e 'NVIDIA_VISIBLE_DEVICES'='all'
  -e 'NVIDIA_DRIVER_CAPABILITIES'='all'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:5000]'
  -l net.unraid.docker.icon='https://raw.githubusercontent.com/yayitazale/unraid-templates/main/frigate.png'
  -p '5000:5000/tcp'
  -v '/mnt/user/appdata/frigate2':'/config':'rw'
  -v '/mnt/user/appdata/frigate2/Media':'/media/frigate':'rw'
  -v '/etc/localtime':'/etc/localtime':'rw'
  --runtime=nvidia
  --shm-size=256mb
  --mount type=tmpfs,target=/tmp/cache,tmpfs-size=1000000000
  --restart unless-stopped 'ghcr.nju.edu.cn/blakeblackshear/frigate:stable-tensorrt'

Relevant Frigate log output

s6-rc: info: service s6rc-fdholder: starting
s6-rc: info: service s6rc-oneshot-runner: starting
s6-rc: info: service s6rc-oneshot-runner successfully started
s6-rc: info: service fix-attrs: starting
s6-rc: info: service s6rc-fdholder successfully started
s6-rc: info: service fix-attrs successfully started
s6-rc: info: service legacy-cont-init: starting
s6-rc: info: service legacy-cont-init successfully started
s6-rc: info: service trt-model-prepare: starting
s6-rc: info: service log-prepare: starting
s6-rc: info: service log-prepare successfully started
s6-rc: info: service nginx-log: starting
s6-rc: info: service go2rtc-log: starting
s6-rc: info: service frigate-log: starting
s6-rc: info: service certsync-log: starting
s6-rc: info: service go2rtc-log successfully started
s6-rc: info: service go2rtc: starting
s6-rc: info: service nginx-log successfully started
s6-rc: info: service frigate-log successfully started
s6-rc: info: service certsync-log successfully started
s6-rc: info: service go2rtc successfully started
s6-rc: info: service go2rtc-healthcheck: starting
s6-rc: info: service go2rtc-healthcheck successfully started
s6-rc: warning: unable to start service trt-model-prepare: command exited 4
Generating the following TRT Models: yolov7-320
Downloading yolo weights
2024-12-01 11:11:46.113449109  [INFO] Preparing new go2rtc config...
2024-12-01 11:11:47.229158848  [INFO] Starting go2rtc...
2024-12-01 11:11:47.327835284  11:11:47.327 INF go2rtc platform=linux/amd64 revision=b2399f3 version=1.9.2
2024-12-01 11:11:47.327839481  11:11:47.327 INF config path=/dev/shm/go2rtc.yaml
2024-12-01 11:11:47.328331173  11:11:47.328 INF [api] listen addr=:1984
2024-12-01 11:11:47.328425433  11:11:47.328 INF [rtsp] listen addr=:8554
2024-12-01 11:11:47.328674906  11:11:47.328 INF [webrtc] listen addr=:8555/tcp
2024-12-01 11:11:55.930287914  [INFO] Starting go2rtc healthcheck service...

Install method

Docker CLI

Object Detector

TensorRT

Screenshots of the Frigate UI's System metrics pages

Can‘t see UI

Any other information that may be helpful

No response

NickM-27 · 2024-12-01T15:05:25Z

NickM-27
Dec 1, 2024
Collaborator Sponsor

In 0.15 models are not downloaded by default and you can download a yolonas model separately to run using ONNX

5 replies

yangrusen Dec 3, 2024
Author

Oh, I don’t have a background in computer science; I can only follow tutorials step by step. I think trying a VPN is the best solution.
Thank you very much for your reply!

NickM-27 Dec 3, 2024
Collaborator Sponsor

it is a pretty easy step by step https://deploy-preview-13787--frigate-docs.netlify.app/configuration/object_detectors#yolo-nas-2

yangrusen Dec 24, 2024
Author

I finally downloaded the Yolonas model and changed config to below. But the container is still trying to download .trt model and stuck. @NickM-27

detectors:

onnx:

type: onnx

model:

model_type: yolonas

width: 320 # <--- should match whatever was set in notebook

height: 320 # <--- should match whatever was set in notebook

input_tensor: nchw

input_pixel_format: bgr

path: /config/yolo_nas_s.onnx

labelmap_path: /labelmap/coco-80.txt

yangrusen Dec 24, 2024
Author

Frigate log:
s6-rc: info: service s6rc-fdholder: starting
s6-rc: info: service s6rc-oneshot-runner: starting
s6-rc: info: service s6rc-oneshot-runner successfully started
s6-rc: info: service fix-attrs: starting
s6-rc: info: service s6rc-fdholder successfully started
s6-rc: info: service fix-attrs successfully started
s6-rc: info: service legacy-cont-init: starting
s6-rc: info: service legacy-cont-init successfully started
s6-rc: info: service trt-model-prepare: starting
s6-rc: info: service log-prepare: starting
s6-rc: info: service log-prepare successfully started
s6-rc: info: service nginx-log: starting
s6-rc: info: service go2rtc-log: starting
s6-rc: info: service frigate-log: starting
s6-rc: info: service certsync-log: starting
s6-rc: info: service nginx-log successfully started
s6-rc: info: service frigate-log successfully started
s6-rc: info: service go2rtc-log successfully started
s6-rc: info: service go2rtc: starting
s6-rc: info: service certsync-log successfully started
s6-rc: info: service go2rtc successfully started
s6-rc: info: service go2rtc-healthcheck: starting
s6-rc: info: service go2rtc-healthcheck successfully started
Generating the following TRT Models: yolov7-320
Downloading yolo weights
2024-12-24 14:54:46.108978495 [INFO] Preparing new go2rtc config...
2024-12-24 14:54:46.471286316 [INFO] Starting go2rtc...
2024-12-24 14:54:46.560816090 14:54:46.560 INF go2rtc platform=linux/amd64 revision=b2399f3 version=1.9.2
2024-12-24 14:54:46.560820415 14:54:46.560 INF config path=/dev/shm/go2rtc.yaml
2024-12-24 14:54:46.561200138 14:54:46.561 INF [rtsp] listen addr=:8554
2024-12-24 14:54:46.561203000 14:54:46.561 INF [api] listen addr=:1984
2024-12-24 14:54:46.561343722 14:54:46.561 INF [webrtc] listen addr=:8555/tcp
2024-12-24 14:54:56.081031195 [INFO] Starting go2rtc healthcheck service...

NickM-27 Dec 24, 2024
Collaborator Sponsor

This is fixed in the latest 0.15 beta

hh-TDA · 2024-12-05T02:57:40Z

hh-TDA
Dec 5, 2024

Hello,
Are you saying that you want to download the model in an environment without network restrictions and then use it in an environment with network restrictions?

2 replies

yangrusen Dec 5, 2024
Author

Actually, that’s not my intention. My Docker container is unable to download the model on its own due to the Great Firewall in China.
I can access the necessary webpages from my company network, but not from my home server, as I currently don’t have a VPN.

By the way, does the YOLO-NAS model generation work for you? I’ve been unable to generate models, even when following the steps one by one. https://deploy-preview-13787--frigate-docs.netlify.app/configuration/object_detectors#yolo-nas-2

hh-TDA Dec 5, 2024

I am training with Darknet YOLO, creating my custom model, and using it.
version : stable-tensorrt

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Detector Support]: Tensorrt version won't start - stuck on starting go2rtc healthcheck service #15288

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 2 comments 7 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

Select a reply

[Detector Support]: Tensorrt version won't start - stuck on starting go2rtc healthcheck service #15288

yangrusen Dec 1, 2024

Describe the problem you are having

Version

Frigate config file

docker-compose file or Docker CLI command

Relevant Frigate log output

Install method

Object Detector

Screenshots of the Frigate UI's System metrics pages

Any other information that may be helpful

Replies: 2 comments · 7 replies

NickM-27 Dec 1, 2024 Collaborator Sponsor

yangrusen Dec 3, 2024 Author

NickM-27 Dec 3, 2024 Collaborator Sponsor

yangrusen Dec 24, 2024 Author

yangrusen Dec 24, 2024 Author

NickM-27 Dec 24, 2024 Collaborator Sponsor

hh-TDA Dec 5, 2024

yangrusen Dec 5, 2024 Author

hh-TDA Dec 5, 2024

yangrusen
Dec 1, 2024

Replies: 2 comments 7 replies

NickM-27
Dec 1, 2024
Collaborator Sponsor

yangrusen Dec 3, 2024
Author

NickM-27 Dec 3, 2024
Collaborator Sponsor

yangrusen Dec 24, 2024
Author

yangrusen Dec 24, 2024
Author

NickM-27 Dec 24, 2024
Collaborator Sponsor

hh-TDA
Dec 5, 2024

yangrusen Dec 5, 2024
Author