Deploy a function inside a specific network

[emdant]

Hello, I have a docker-compose file in which I run a container running nuctl, which then deploys some functions. The container which are then deployed won't join the docker-compose network, making difficult to communicate between the services inside the network and the functions.

Is there any way to deploy the function inside a network, or some way to bypass this problem?

[liranbg]

on your function configuration, under platform.attributes block you may use network and specify the docker network you want your function to use.

e.g.:

meta:
  name: my-func
spec:
  handler: main:handler
  runtime: python:3.8
  platform:
    attributes:
      network: my-docker-network

[emdant]

Even by specifying the network in platform.attributes, when I inspect the container it is connected only to the default bridge, not on the created network

[efcy]

I have the same problem. I can't get this to work. Even if a disable the default bridge network in the daemon.json.

Also I could not find this network setting in the documentation.

[efcy]

This is related to https://github.com/nuclio/nuclio/issues/989 So i guess it worked for someone at some point.

[efcy]

@emdant I made it work. I had platform twice in my file. In this case it does not work correctly.

i added the name (cvat_cvat ) I got from docker network ls:

NETWORK ID     NAME        DRIVER    SCOPE
30fef26c5013   cvat_cvat   bridge    local

to the yaml file:

metadata:
  name: dummy-model
  namespace: cvat
  annotations:
    name: Dummy Model
    type: detector
    framework: pytorch
    spec: |
      [
        { "id": 1, "name": "ball" }
      ]

spec:
  description: Detectron2
  runtime: 'python:3.8'
  handler: main:handler
  eventTimeout: 30s

  build:
    image: cvat/dummy
    baseImage: ubuntu:20.04

    directives:
      preCopy:
        - kind: ENV
          value: DEBIAN_FRONTEND=noninteractive
        - kind: RUN
          value: apt-get update && apt-get -y install curl git python3 python3-pip libgl1-mesa-glx libglib2.0-0 python-is-python3
        - kind: RUN
          value: python3 -m pip install opencv-python

  triggers:
    myHttpTrigger:
      maxWorkers: 2
      kind: 'http'
      workerAvailabilityTimeoutMilliseconds: 10000
      attributes:
        maxRequestBodySize: 33554432 # 32MB

  platform:
    attributes:
      restartPolicy:
        name: always
        maximumRetryCount: 3
      mountMode: volume
      network: cvat_cvat

Maybe my example can help you.