"docker: command not found" error when running code-deploy pipe

[Julien Richard-Foy]

I’ve copied the `script` entry given in the documentation of the atlassian/aws-code-deploy pipe and adapted to my needs:

pipelines:
  branches:
    master:
      - step:
        name: Build
        image: node:10
        script:
          - ...
      - step:
        name: Deploy
        script:
          - pipe: atlassian/aws-code-deploy:0.2.1
            variables:
              AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID
              AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY
              AWS_DEFAULT_REGION: 'us-east-2'
              APPLICATION_NAME: $APPLICATION_NAME
              COMMAND: 'upload'
              ZIP_FILE: $ZIP_FILE
              S3_BUCKET: $S3_BUCKET

When the pipeline is run, I get the following error:

 docker container run \
    --volume=/opt/atlassian/pipelines/agent/build:/opt/atlassian/pipelines/agent/build \
    --volume=/opt/atlassian/pipelines/agent/ssh:/opt/atlassian/pipelines/agent/ssh:ro \
      ...
bash: docker: command not found

My pipeline is using the default Docker image.

[Julien Richard-Foy · Accepted solution]

I was able to fix the problem by adding the `docker` service to my step:

 - step:
   name: Deploy
   services:
     - docker
   script:
     - pipe: atlassian/aws-code-deploy:0.2.1
       variables:
         AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID
         ...

[Radu Cristescu]

For me, this was caused by BCLOUD-22851. I have self-hosted runners where this error happens once in a blue moon.

I have a cron job on the lookout for this bug. For each runner, it looks for the `docker` directory. If it finds it, it stops the runner, removes the directory, and starts the runner again.

Unfortunately, a pipeline has to trigger the problem before the workaround kicks in, so the pipeline that triggers it will fail. But because the workaround kicks in automatically via cron, rerunning the pipeline will be fine.

[helgispbru]

I have the same issue and this solutions doesn't help me.

[Dipesh Rane]

pipelines:
  branches:
    master:
      - step:
        name: Build
        image: node:10
        script:
          - ...
      - step:
        services:
         - docker
        name: Deploy
        script:
          - pipe: atlassian/aws-code-deploy:0.2.1
            variables:
              AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID
              AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY
              AWS_DEFAULT_REGION: 'us-east-2'
              APPLICATION_NAME: $APPLICATION_NAME
              COMMAND: 'upload'
              ZIP_FILE: $ZIP_FILE
              S3_BUCKET: $S3_BUCKET

[Dipesh Rane]

add docker as services it will fix the issue

[Peter]

You must have a reserved Variable called PATH, in your enviroment. 

[Xinyi Xu]

Hey, accepted answer doesn't work for me. I was able to make it run adding optioins:docker: true to the end. See below

pipelines:
  branches:
    master:
      - step:
        name: Build
        image: node:10
        script:
          - ...
      - step:
        name: Deploy
        script:
          - pipe: atlassian/aws-code-deploy:0.2.1
            variables:
              AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID
              AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY
              AWS_DEFAULT_REGION: 'us-east-2'
              APPLICATION_NAME: $APPLICATION_NAME
              COMMAND: 'upload'
              ZIP_FILE: $ZIP_FILE
              S3_BUCKET: $S3_BUCKET

options:
  docker: true

[Samuell]

I had same issue i just reruned pipelines again and it works.