As part of running the pipe below
- pipe: atlassian/aws-lambda-deploy:0.5.7
variables:
AWS_ACCESS_KEY_ID: ${AWS_ACCESS_KEY_ID}
AWS_SECRET_ACCESS_KEY: ${AWS_SECRET_ACCESS_KEY}
AWS_DEFAULT_REGION: 'eu-west-2'
FUNCTION_NAME: 'MyFunction'
COMMAND: 'update'
ZIP_FILE: 'code.zip'
The script update-lambda.sh writes to /opt/atlassian/pipelines/agent/build/.bitbucket/pipelines/generated/pipeline/pipes/aws-lambda-deploy-env. That file is then read by pipe.sh in a subsequent pipe when you run with the alias command. e.g.
COMMAND: 'alias'
However if I have another pipe in my bitbucket-pipelines.yml file before the aws-lambda-deploy pipe. e.g.
sonarsource/sonarcloud-scan:1.2.1
then update-lamda.sh cannot write the aws-lambda-deploy-env file as it gets a permission denied error.
Shouldn't the aws-lambda-deploy-env file be created in /opt/atlassian/pipelines/agent/build/.bitbucket/pipelines/generated/pipeline/pipes/atlassian/aws-lambda-deploy ?
And why does the creation of the /opt/atlassian/pipelines/agent/build/.bitbucket/pipelines/generated/pipeline/pipes/ by another pipe stop it being available for the subsequent pipe?