Use the Variable in definitions

[Mohan Kona]

I'm trying to send variable to image in services, bit it seems like the variable is not being accepted and it throws error in the line where image with variable is defined.

Please find the yaml file below

image: image:latest definitions: steps: - step: &udpateEnvVariable name: Populate Environment script: - export ENV=dev - if [ "$FRONTEND_IMAGE" = "release" ]; then export ENV=qa; elif [ "$FRONTEND_IMAGE" = "daily" ]; then export ENV=dev; else export ENV=dev; fi services: frontend: image: name: image_name:$IMAGE options: docker: true pipelines: custom: dummy: - variables: - name: IMAGE default: daily allowed-values: - daily - release - misc - step: *udpateEnvVariable - step: name: Dummy services: - frontend script: - echo "ENV is $ENV" 

 and below is the error thrown

[Aron Gombas _Midori_ · Accepted solution]

I don't think that the YAML above is valid at all. Validate this with an online validator before anything else.

YAML is very sensitive to indentation and line-breaks, those have important meaning.

[Mohan Kona]

I couldn't paste the code in yaml format, but to make it simple, i'm trying to provide service value as a variable, something like

definitions: 
 services:
  frontend:
   image:
    name: $docker_image

 OR

services:
  - $docker_image

[Aron Gombas _Midori_]

The variable in this case is a regular Bitbucket pipeline variable?

Also, I don't know, but I guess it is rather likely that variable names are case-sensitive.

[Mohan Kona]

no, the variable here is the user defined one.

[Patrik S]

Hey @Mohan Kona ,

Just adding to @Aron Gombas _Midori_ response, using variables to define image names and services is not yet possible directly in the YAML file.

However, Bitbucket Cloud has recently launched a new feature called Dynamic Pipelines. This feature is designed to make every part of the Pipeline workflow easy to customize and adjust with parameters/variables.

With Dynamic Pipelines, you can use Variable and Deployment APIs to add any custom variables you need at any point in the Pipeline.

You can find more information on how to use these APIs here: Atlassian Community Link.

Thank you, @Mohan Kona !

Patrik S