Use pipline to update layer

Question

Hello,

I am using bitbucket pipline to deploy my lambdas. For now everything is fine, except for deploying layers.

in 2018 aws introduced layer for lambdas, mainly used to share a common code (conf, models, ...) between lambdas and also to create runtime environement for not supported langages.

I have lambdas to manage users (nodejs + mongodb) : create user, login, activate user, update, ...

these lambdas are sharing the same model and configuration. So, I would like to avoid duplicating code, so I thought about creating layer.

My problem is:

1/ How can I do to update my layer when I deploy my lambda (specify latest layer) ?

I am using (AWS Lambda deploy : v0.3.2maintained by Atlassian)

Thank you in advance for your help.

Best regards

Adel MANI

Yadvendra Singh · Answer

Hi @Alexander Zhukov I tried the same config but failed. just removed the account number for security. here is my config file ---------------------------- {"Layers": [ { "Arn": "arn:aws:lambda:eu-west-1:xxxxxxxxxxx:layer:ap-mail-node-pgpool-1:22", "CodeSize": 229879 } ] } --------------------------------------- function update part works just file but while updating config it fails here is the error log INFO: Updating Lambda function configuration.aws lambda update-function-configuration --function-name ap-mail-ds1-scs-1 --cli-input-json {"Layers": [ { "Arn": "arn:aws:lambda:eu-west-1:xxxxxxxxxxx:layer:ap-mail-node-pgpool-1:22", "CodeSize": 229879 } ]}Parameter validation failed:Invalid type for parameter Layers[0], value: {u'CodeSize': 229879, u'Arn': u'arn:aws:lambda:eu-west-1:xxxxxxxxxxx:layer:ap-mail-node-pgpool-1:22'}, type: , valid types: ✖ Failed to update Lambda function configuration.

MANI Adel · Answer

Hi @Alexander Zhukov, 
Thank you for replying my request.

My question was not clair, I was looking for the way how to deploy an update or publish of layer (publish-layer-version).

I solved this problem by creating aws_nodejs docker image and lauching the the following CLI command :

script:  
    - aws configure set default.region $AWS_DEFAULT_REGION 
    - aws configure set aws_access_key_id $AWS_ACCESS_KEY_ID 
    - aws configure set aws_secret_access_key $AWS_SECRET_ACCESS_KEY 
    - aws lambda publish-layer-version --layer-name Layer_01 --compatible-runtimes nodejs nodejs10.x nodejs8.10 --zip-file fileb://nodejs.zip --region $AWS_DEFAULT_REGION

Thank you very much for taking in consideration my request.

Best regards

Adel

Alexander Zhukov_1bqyxb · Answer

Hi @MANI Adel . If you wan't to just add an existing layer to your lambda function, you should create a function configuration json file an reference it in the FUNCTION_CONFIGURATION variable. The pipe will update the function configuration to include the layers.

Example function-configuration.json

{"Layers": 
  [ 
    { "Arn": "arn:aws:lambda:us-east-2:123456789012:layer:my-layer:3", "CodeSize": 169 }, 
    { "Arn": "arn:aws:lambda:us-east-2:210987654321:layer:their-layer:2", "CodeSize": 169 } 
  ] 
}

Example pipe configuration:

script:
- pipe: atlassian/aws-lambda-deploy:0.4.1
variables:
AWS_ACCESS_KEY_ID: $AWS_ACCESS_KEY_ID
AWS_SECRET_ACCESS_KEY: $AWS_SECRET_ACCESS_KEY
AWS_DEFAULT_REGION: $AWS_DEFAULT_REGION
FUNCTION_NAME: 'my-lambda-function'
COMMAND: 'update'
ZIP_FILE: 'my-function-code.zip'
      FUNCTION_CONFIGURATION: 'function-configuration.json'

Note, that updating or creating a function layer is not supported in the pipe yet.