onvif-camera-mocking

This project started as a fork from kate-goldenring/onvif-camera-mocking This project consists of tools and instructions for mocking an ONVIF-compliant IP camera and passing an RTSP stream to simulate a live video stream.

If you want to use the container without building it, you can download it using the following command:

docker pull winiotsaleskit.azurecr.io/onvif-camera-mocking:latest

Steps

Note: these steps only work on Linux and have only been tested on Ubuntu 22.04LTS

1. Set up your developer machine

Install building dependencies

sudo apt update
sudo apt install flex bison byacc make m4 autoconf unzip \
    git g++ wget -y

Install gstreamer dependencies

sudo apt update
sudo apt install gstreamer-1.0 libgstrtspserver-1.0-dev gstreamer1.0-rtsp \
    gstreamer1.0-plugins-ugly python3-gi

Clone the repository

git clone https://github.com/fcabrera23/onvif-camera-mock

2. Build the ONVIF server

The onvif_srvd project was forked, and the updated original dependencies to keep it up to date with gsoap version and fix broken dependencies.

Build the onvif_srvd
```
cd onvif_srvd
make release
```

Ensure the build was created - Using the ls command, check the onvif_srvd is available. You should see something like this:

ubuntu@NUC-Ubuntu:/home/ubuntu/onvif-camera-mocking/onvif_srvd# ls
generated  gsoap-2.8  LICENSE  Makefile  onvif_srvd  README.md  SDK  src  start_scripts  wsdl

3. Build the WS-Discovery Service

The wsdd project was forked, and the updated original dependencies to keep it up to date with gsoap version and fix broken dependencies.

Move to the project root folder and build the wsdd
```
cd wssd
make release
```

Ensure the build was created - Using the ls command, check the used is available. You should see something like this:

ubuntu@NUC-Ubuntu:/home/ubuntu/onvif-camera-mocking/wsdd# ls
CHANGELOG.md  generated soap-2.8 LICENSE  Makefile  README.md  SDK  src  start_scripts  wsdd  wsdl

4. Start the ONVIF and Discovery services with the RTSP feed

Run ifconfig or ipconfig to determine your network interface. Then, pass your interface (such as eno1,eth0, eth1, etc) to the script. The following assumes eth0.

Run the main.py python script to start the discovery services and RTSP feed. You can specify the network interface and optionally the resources directory, firmware version of the camera, and path to the MP4 streaming video file. The script uses the following environment variables and defaults:

Argument	Defaults	Description
INTERFACE	No - Mandatory	Network interface to expose discovery service
DIRECTORY	Defaults to `$PWD`	The directory of the project
FIRMWARE	Default to 1.0	The "mock" firmware version of the camera
MP4FILE	Default to mocking color stripe video	sPath to the video mp4 location

Run the mocking camera using the following command:

export INTERFACE=<interface>
python3 main.py

5. Ensure that the ONVIF camera service is running and discoverable

Use one of the tools recommended by onvif_srvd for testing the ONVIF service.

Program	OS
ONVIF Device Manager	Windows
ONVIF Device Tool (GUI)	Linux
gsoap-onvif	Linux

Run the tool on the same device or network as your newly mocked camera. In the tool, look for a new camera called TestDev.

6. Build Docker container

You can build this as a container using the [Dockerfile.onvif-camera]. Note: Building the container can take up to 20-30 minutes. Once built, make sure to run it with the appropriate environment variables mentioned in Section 4.

7. Run it with Akri and Kubernetes

If you are using K3s on a Linux device, please follow the ONVIF for IP cameras guide.

Once you have your Kubernetes cluster running and Akri installed, run the following command to enable the ONVIF discovery:
```
sudo ip route add 239.255.255.250/32 dev cni0
```
Note: If you're not using Flannel, you may need to check the Kubernetes network interface name and change it above.
1. Find the IP of the mock ONVIF container:
```
kubectl get pods -o wide
```
1. Run route and find the network interface name that matches the IP of the ONVIF pod (i.e. cali909b8c65537).
2. Now enable the ONVIF discovery:
```
sudo ip route add 239.255.255.250/32 dev <insert interface name>
```

If you are using AKS Edge Essentials, please follow the Discover ONVIF cameras with Akri guide.

Once you have your Kubernetes cluster running and Akri installed, run the following PowerShell command to enable the ONVIF discovery:

Invoke-AksEdgeNodeCommand -command "sudo ip route add 239.255.255.250/32 dev cni0"

Note: If you're not using Flannel, you may need to check the Kubernetes network interface name and change it above.

Find the IP of the mock ONVIF container:

kubectl get pods -o wide

2.Find the network interface name that matches the IP of the ONVIF pod (i.e. cali909b8c65537):

Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "route"

Now enable the ONVIF discovery:

Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "sudo ip route add 239.255.255.250/32 dev <insert interface name>"

If you're using Scalable-cluster with an External virtual switch, you will need to enable the --dport 3702 and --sport 3702 firewall rule using the following command:

Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "sudo iptables -A INPUT -p udp --dport 3702 -j ACCEPT"
Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "sudo sed -i '/-A OUTPUT -j ACCEPT/i-A INPUT -p udp -m udp --dport 3702 -j ACCEPT' /etc/systemd/scripts/ip4save"   
Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "sudo iptables -A INPUT -p udp --sport 3702 -j ACCEPT"
Invoke-AksEdgeNodeCommand -NodeType "Linux" -command "sudo sed -i '/-A OUTPUT -j ACCEPT/i-A INPUT -p udp -m udp --sport 3702 -j ACCEPT' /etc/systemd/scripts/ip4save"

If you everything was correctly set up, you should see the mocked camera discovered by Akri.

PS C:\Users\Administrator\Desktop> kubectl get akrii
NAME                CONFIG       SHARED   NODES             AGE
akri-onvif-c9582f   akri-onvif   true     ["node1-ledge"]   7m28s
PS C:\Users\Administrator\Desktop>

8. Using a custom video RTSP feed

Finally, if you want to create an RTSP feed with a custom video, copy the video file into the AKS-EE host first. For example, if you want to use the sample.mp4 file from your directory C:\Users\Admin\sample.mp4, copy it to the AKS-EE Linux node /home/aksedge-user/sample.mp4 file using the following cmd:

Copy-AksEdgeNodeFile -FromFile C:\Users\fcabrera\Downloads\sample-15s.mp4 -toFile /home/aksedge-user/sample.mp4 -PushFile

For more information, see Copy-AksEdgeNodeFile.

Last step is to modify your Pod deployment to mount the necessary file and use that file in the RTSP streamer server. The following example uses the public docker container and mounts the file sample.mp4 to create the RSTP stream, but you can change the deployment for your environment.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: onvif-camera-mocking
spec:
  replicas: 1
  selector:
    matchLabels:
      app: onvif-camera-mocking
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 1
  minReadySeconds: 5    
  template:
    metadata:
      labels:
        app: onvif-camera-mocking
    spec:
      nodeSelector:
        "kubernetes.io/os": linux
      containers:
      - name: azure-vote-front
        image: winiotsaleskit.azurecr.io/onvif-camera-mocking:latest
        ports:
        - containerPort: 8554
        - containerPort: 1000
        - containerPort: 3702
        env:
        - name: INTERFACE
          value: "eth0"
        - name: DIRECTORY
          value: "/onvif-camera-mock"
        - name: MP4FILE
          value: /mnt/sample.mp4
        volumeMounts:
        - name: sample-volume
          mountPath: /mnt
      volumes:
      - name: sample-volume
        hostPath:
          path: /home/aksedge-user
          type: Directory

fcabrera23 / onvif-camera-mock