Issue Summary

If bitbucket-pipelines.yml is configured to run a large number of pipes in a single build, Pipelines will exert the following behaviours:

1. Received "An unidentified error occurred" while running Pipelines through UI
2. Multiple builds of the same commit triggered after a single push to the branch

This greatly impacts user who has a large number of AWS lambda function and relies on Pipe to deploy it.

Steps to Reproduce

Duplicate the following pipes 190 times in the bitbucket-pipelines.yml config:

pipelines:
  default:
    - step:
        script:
          - pipe: docker://bitbucketpipelines/aws-lambda-deploy:0.5.0
             variables:
                 AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID_STAGE
                 AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY_STAGE
                 AWS_DEFAULT_REGION: $AWS_REGION_STAGE
                 FUNCTION_NAME: my_function_name
                 COMMAND: "update"
          - pipe: docker://bitbucketpipelines/aws-lambda-deploy:0.5.0
             variables:
                 AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID_STAGE
                 AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY_STAGE
                 AWS_DEFAULT_REGION: $AWS_REGION_STAGE
                 FUNCTION_NAME: my_function_name
                 COMMAND: "update"
.....
.....

Ideally, the "FUNCTION_NAME" will be different in different pipes, but we're able to reproduce with the same config as above

Expected Results

• Users able to run build through UI without error
• Pipelines only trigger a single build after a push is performed.

Actual Results

1. Received "An unidentified error occurred" while running Pipelines through UI
2. Multiple builds triggered on the same commit, after a single push to the branch

Workaround

Two workarounds:

1. Reduce the number of Pipes within a single build
   • If you need to trigger more than 190 pipes, you can configure the remaining Pipes to run in subsequence build with Pipelines custom step and trigger it with Trigger Pipelines Pipe
     Example:

     pipelines:
       default:
         - step:
             script:
               - pipe: docker://bitbucketpipelines/aws-lambda-deploy:0.5.0
                  variables:
     ......
         - step:
             name: Trigger subsequent pipeline with custom step
             script:
               - pipe: atlassian/trigger-pipeline:5.0.0
                  variables:
                       BITBUCKET_USERNAME: '<string>'
                       BITBUCKET_APP_PASSWORD: '<string>'
                       REPOSITORY: '<string>'
                       ACCOUNT: '<string>' # Optional
                       REF_TYPE: 'branch' # Optional
                       REF_NAME: 'stage' # Optional
                       CUSTOM_PIPELINE_NAME: 'SubsequentStepswithpipes' 
                       # PIPELINE_VARIABLES: '<json>' # Optional
                       # WAIT: '<boolean>' # Optional
                       # WAIT_MAX_TIMEOUT: '<string>' # Optional
                       # DEBUG: '<boolean>' # Optional
         custom : 
           SubsequentStepswithpipes : 
            - step:
                script : 
                - pipe: atlassian/aws-ecr-push-image:1.1.1
     .....
     

2. If it's an Atlassian Pipes(pipe: atlassian/****), you can replace the Pipe name with

   pipe: docker://bitbucketpipelines/*****
   

   For example: pipe: docker://bitbucketpipelines/aws-lambda-deploy:0.5.0