I have a deployment that is deploying based on the stage sls deploy --stage staging and I want to export the lambda ARN from that staging function within the Outputs. I expected it to understand the Condition but it seems it doesn't and is trying to deploy to each environment regardless of the Condition.
Is there a way to do it?
service: test
frameworkVersion: '^2.2.0'
plugins:
...
provider:
name: aws
runtime: python3.8
stage: ${opt:stage}
resources:
Conditions:
IsStaging:
Fn::Equals:
- ${self:provider.stage}
- staging
Resources:
...
Outputs:
MyLambdaFunction:
Condition: IsStaging
Value: !GetAtt [ MyLambdaFunction, Arn ]
Export:
Name: MyLambdaFunction
I couldn't find a plugin that mention the Conditional Output, is it something that works in a more up to date framework version or is it just not working?
Related
I'm trying to define my serverless framework deployment bucket.
My serverless.yml looks like this:
provider:
name: aws
runtime: nodejs14.x
region: us-east-1
stage: dev
deploymentBucket:
name: ${self:environment.DEPLOYMENT_BUCKET}
environment:
${file(../evn.${opt:stage, 'dev'}.json)}
and the evn.dev.json file looks like this:
{
"DEPLOYMENT_BUCKET": "myBucketName"
}
(both of these files have non-relevant parts removed)
I'm getting a "cannot resolve variable at "provicer.deploymentBucket.name" error when trying to deploy.
How do I reference the DEPLOYMENT_BUCKET variable in the serverless.yml file?
EDIT: Other errors:
${environment}:DEPLOYMENT_BUCKET -> Could not locate deployment bucket. Error: The specified bucket is not valid
name: ${environment:DEPLOYMENT_BUCKET}1 -> Unrecognized configuration variable sources: "environment"
name: ${self:provider.environment:DEPLOYMENT_BUCKET}
and
name: ${self:environment:DEPLOYMENT_BUCKET}
-> Cannot resolve serverless.yml: Variables resolution errored with - Cannot resolve variable at "provider.deploymentBucket.name": Value not found at "self" source
I was able to solve the problem with this:
${file(../evn.${opt:stage, 'dev'}.json):DEPLOYMENT_BUCKET}
But 'reading' that file twice -- both here and in the 'environment' area seems to somewhat defeat the purpose of the environments area.
I have a serverless.common.yml, with properties that should be shared by all the services, with that:
service: ixxxx
custom:
stage: ${opt:stage, self:provider.stage}
resourcesStages:
prod: prod
dev: dev
resourcesStage: ${self:custom.resourcesStages.${self:custom.stage}, self:custom.resourcesStages.dev}
lambdaPolicyXRay:
Effect: Allow
Action:
- xray:PutTraceSegments
- xray:PutTelemetryRecords
Resource: "*"
And, another serverless.yml inside a services folder, which uses properties on the common file:
...
custom: ${file(../../serverless.common.yml):custom}
...
environment:
stage: ${self:custom.stage}
...
In that way, I can access the custom variables (from the common file) without a problem.
Now, I want to continue to import this file to custom, but adding new variables, related to this service, to it, so I tried that:
custom:
common: ${file(../../serverless.common.yml):custom}
wsgi:
app: app.app
packRequirements: false
pythonRequirements:
dockerizePip: non-linux
And it seems it's possible to access, for example:
environment:
stage: ${self:custom.common.stage}
But now, I'm receiving the error:
Serverless Warning --------------------------------------
A valid service attribute to satisfy the declaration 'self:custom.stage' could not be found.
Serverless Warning --------------------------------------
A valid service attribute to satisfy the declaration 'self:custom.stage' could not be found.
Serverless Error ---------------------------------------
Trying to populate non string value into a string for variable ${self:custom.stage}. Please make sure the value of the property is a strin
What am I doing wrong?
In your serverless.common.yml you must reference as if it were serverless.yml. In this case ${self:custom.stage} does not exist, but ${self:custom.common.stage} does exist.
service: ixxxx
custom:
stage: ${opt:stage, self:provider.stage}
resourcesStages:
prod: prod
dev: dev
resourcesStage: ${self:custom.common.resourcesStages.${self:custom.common.stage}, self:custom.resourcesStages.dev}
lambdaPolicyXRay:
Effect: Allow
Action:
- xray:PutTraceSegments
- xray:PutTelemetryRecords
Resource: "*"
So I'm pretty new to CloudFormation and also to Serverless framework. I've been trying to work through some exercises (such as an automatic thumbnail generator) and then create some simple projects that I can hopefully generalize for my own purposes.
Right now I'm attempting create a stack/function that creates two S3 buckets and has the Lambda Function take a CSV file form one, perform some simple transformations, and place it in the other receiving bucket.
In trying to build off the exercise I've done, I created a Yaml file with the following code:
provider:
name: aws
runtime: python3.8
region: us-east-1
profile: serverless-admin
timeout: 10
memorySize: 128
iamRoleStatements:
- Effect: "Allow"
Action:
- "s3:*"
Resource: "*"
custom:
assets:
targets:
- bucket1: csvbucket1-08-16-2020
pythonRequirements:
dockerizePip: true
- bucket2: csvbucket2-08-16-2020
pythonRequirements:
dockerizePip: true
functions:
protomodel-readcsv:
handler: handler.readindata
events:
s3:
- bucket: ${self:custom.bucket1}
event: s3:ObjectCreated:*
suffix: .csv
- bucket: ${self:custom.bucket2}
plugins:
- serverless-python-requirements
- serverless-s3-deploy
However, when i do a Serverless deploy from my command prompt, I get:
Serverless Warning --------------------------------------
A valid service attribute to satisfy the declaration 'self:custom.bucket1' could not be found.
Serverless Warning --------------------------------------
A valid service attribute to satisfy the declaration 'self:custom.bucket2' could not be found.
Serverless Error ---------------------------------------
Events for "protomodel-readcsv" must be an array, not an object
I've tried to make the events object in the protohandler-readcsv: by adding a - but I then get a bad indentation error that for some reason I cannot reconcile. But, more fundamentally, I'm not exactly sure why that item would need be an array anyway, and I wasn't clear about the warnings with the buckets either.
So sorry about a pretty newbie question about this, but running tutorials/examples online leaves a lot to try to figure out in trying to generalize/customize these examples.
custom:
assets:
targets:
- bucket1
I guess you need self:custom.assets.targets.bucket1, not sure if this nested assets will work.
Please check the example below is supposed to work.
service: MyService
custom:
deploymentBucket: s3_my_bucket
provider:
name: aws
deploymentBucket: ${self:custom.deploymentBucket}
stage: dev
In Serverless.yml, I defined resource:
provider:
name: aws
runtime: nodejs6.10
region: us-east-1
stage: dev
environment:
customerDef: myvariable
resources:
Resources:
NewResource:
Type: AWS::S3::Bucket
Properties:
BucketName: ${self:service.name}-${self:provider.stage}-uploads
while in handler.js file which is write handle function.
How to get the reference of BucketName?
How to get the Bucket URI?
How to get the customerDef variable value? (provider->environment->customerDef)
All the environment variables defined under the environment node are available at any .js file using process.env.<variable_name>.
In your case, to access the customerDef variable you should use process.env.customerDef.
You can do the same with the BucketName and Bucket URI.
If you have your variables in environment key, you can reference them by process.env.yourVariable in every part of your project
I have the following semver setup:
- name: version
type: semver
source:
driver: gcs
bucket: my-ci
json_key: ((my.serviceaccount))
key: version/version.txt
initial_version: 0.0.0
In my publishjob, I have the following:
name: publish
serial_groups: [version]
plan:
- get: version
passed: [build]
trigger: true
So, basically the publish job is triggered after build job is passed (version updated)
Now, in the publish job I am creating a docker image and pushing it to gcr.
- put: my-gcr
params:
additional_tags: my/ci/tags
build: mycode
get_params: {skip_download: true}
Here, the image is correctly tagged based on the values in the tags file. However, I want to set these values dynamically based on the current version which can be retreived following this:
https://concoursetutorial.com/miscellaneous/versions-and-buildnumbers/#display-version
How can I use this version number to tag my docker image?
I solved it using the following code:
- put: artifacts
params:
additional_tags: version/number
build: mycode
get_params: {skip_download: true}