I have a CloudFormation master stack. I want to write a tool that allows me to add Lambda functions to the master stack using Boto3.
In order to add the function to the stack I need to be able to
Get Outputs from the master stack to use in the function's template.
Add the Function to the master stack.
I have only been able to get this to work with:
Build, zip, and upload the function to S3
Add the function's template to the master stack's templates. (requires editing the master stack's files)
Deploy the master stack.
I would like to be able to create the function without editing the master stack's files.
(i.e. boto3.get_stack_id -> boto3.add_resource_to_stack_by_stack_id)
Is this possible? If so, how do I do it?
No that is not possible. When updating a stack, you always have to provide a URL for the new stack template, or provide the full template body as a string, or use the previous template.
Source: https://docs.aws.amazon.com/AWSCloudFormation/latest/APIReference/API_UpdateStack.html
There is no API call that allows you to directly add a resource to a stack.
Related
I have a nested Cloud Formation Template (multiple templates within a root template )to create a complete web application.
Lambda is created in the first template and few environment variables are added to it.
The later part of the templates also produces some values that has to be added as environment variables.
Is there a way to attach these environment variables to the existing lambda function?
I don't think so, but there are a few options. If you could change the stack dependency order, you could build the stack creating the values depended upon first. If you cannot, you can store your environment variables in SSM Parameter Store as mentioned in this knowledge center article.
So you set the environment variable to a path where the value can be expected, then when creating the stack that knows the value, you store it at that path. When the lambda runs, you just do get parameter.
Is there anyway to make CloudFormation parameter dynamic? I know about the System Manager Parameter, but again I have to change its value manually. I want to use somehow the result of the API call or script(Bash, python) in my CloudFormation resources
for example, as part of the parameter, run a API call to get back some data (any data) and then use/reference the result into the resources, and all in one template.
You can use Cloudformation Custom resource to achieve similar effect, with some caveats.
As an example we can use AWS CDK, which provides a module to create custom resources, and even has a wrapper specifically designed to call AWS API and return the results: https://docs.aws.amazon.com/cdk/api/latest/docs/#aws-cdk_custom-resources.AwsSdkCall.html
Some things to remember:
Custom resource needs to return value in form {'PhysicalResourceId': ..., Data: {"MyAttribute": ...}} in order to support using !GetAtt MyResource.MyAttribute style of reference
Like any other CF resource, Custom resource is not triggered on every update, only if one of the parameters of the resource has changed. So if you supplied some parameter to your API call on stack creation, unless you change value, no update will happen and API call will not be triggered.
I have a single stack with two instances RDS,EC2.I have to follow the structure that has root stack and two nested stack RDS,EC2.
Following should be the structure
1.Root stack
2.Webserver stack: All services related to EC2
3.DB stack: All services related to RDS
The general procedure would be as follows:
Extract rds and ec2 resources to its own template files (e.g. ec2.yaml, rds.yaml)
Parameterize ec2.yaml and rds.yaml, as well as add Outputs section to them. The Outputs are needed if you want to reference their return values in the Root stack.
Upload the ec2.yaml and rds.yaml into S3.
In the root stack create two resources of type AWS::CloudFormation::Stack, i.e. one for rds and one for ec2. Use TemplateUrl to provide paths to locations in S3 (step 3). Use Parameters to define input parameters created in step 2.
Use GetAtt to reference outputs of the nested stack in the root stack.
I am running a AWS CloudFormation stack that takes in some parameters and launches EC2 instances along with other AWS resources. The parameters are fed into the user data of the EC2 instance and based on that changes are made dynamically to the web application residing on the EC2 instance.
UserData:
Fn::Base64:
Fn::Join:
- ""
-
- "#!/bin/bash \n"
- "sh website-conf/website_mysql_config.sh "
- " -c \""
-
Ref: "CompanyName"
As shown in the example above, CompanyName is one of the many parameters passed to the userdata script. The problem is, when any one or multiple of parameters are updated, CloudFormation does not detect that and instead throws this error.
So, in order to update the stack, I have to edit the stack and make changes to the ASG so that CloudFormation 'sees' the changes and executes the stack update.
Is there a way to force CFN to update the stack when the parameters are updated?
CloudFormation will not update the stack unless there is a change in properties of the resources already created in the stack.
For example:
Consider I have a simple template to create a database where I need to pass 2 parameters:
db-name
region
Assume that I am using db-name passing it as value to DBInstanceIdentifier.
Also assume that I am not using the input parameter region for any purpose in creation of resources (or its properties) of the stack in any way.It is more of a dummy parameter I keep for readability purpose.
I passed (TEST-DB1, us-east-1) as input parameters to the CloudFormation template and successfully created the resources.
Scenario-1:
Now if I update the stack(still using the existing template) and just change the input parameters to (TEST-DB2, us-east-1). ie: changing just the db-name and not the region. Then CloudFormation will detect that, this parameter update, results in change in properties of running resource(s) of the stack and will compute and display the modifications as a change set.
Scenario-2:
Suppose I make another update(still using the existing template) property and just change the input parameters to (TEST-DB1, us-east-2). ie: changing just the region and not the db-name. Then CloudFormation will detect that, this parameter update, result in NO change in properties of running resource(s) of the stack will show the Error creating change set.
Bottomline:
Your change in input parameter must result in an update/replacement of any resources(or its attributes like security-groups,port etc..) of the stack. Then AWS CloudFormation will display them as Change Sets for your review. Also, the method (update or replacement) AWS CloudFormation uses depends on which property you update for a given resource type.
Your parameter "CompanyName" is not making any changes to the running
resources of the stack. Hence it is reporting as Error creating
change set. You need to use it to create any resource/resource properties of the stack. Then CloudFormation will detect the change-sets when you modify it. The same applies for any other input-parameters which you use.
Use the AWS CLI Update-Stack command. If you use the AWS CLI you can inject parameters into your stack so any change to any of the parameters result in a new stack. I do this myself to inject the Git/version commit ID into UserData so simply committing changes to the stack's JSON/Yaml to Git will allow stack updates. Any change to the parameters file will allow stack updates, even just a comment. I reference my Git commit ID in UserData the same way you are referencing Ref:CompanyName so when I change the Git commit ID the userData section is updated on stack updates.
Update Stack Command
aws cloudformation update-stack --stack-name MyStack --template-body file:///Users/Documents/Git/project/cloudformation/stack.json --parameters file:///Users/Documents/Git/project/cloudformation/parameters/stack-parameters.dev.json --capabilities CAPABILITY_IAM
Process
With this approach you make your parameters changes to the parameters json or yaml file then check it into version control. Now if you use a build server you can update your stack by checking out master and just running that one line above. Using AWS CodeBuild makes this easy so you don't need jenkins.
The answer of your problem is already answered with this state, CloudFormation will not update the stack unless there is a change in properties of the resources already created in the stack.
And for the answer for your question, please check the explanation below.
There is a way to force Cloudformation to update the stack using the AWS::CloudFormation::Init.
By using cfn-init, each instance can update itself when it detect the change that made by AWS::CloudFormation::Init in metadata.
There is a concept that we must understand first, that is the difference between UserData and metadata, at least under the AWS::CloudFormation::Init case.
Userdata: Will be only called once when the instance is being launch for the first time (this including update that need the instance to be replaced). So, if you update the stack (not creating a new one), even if you change the parameter value, it won't change anything if you call the parameter under UserData.
Metadata: Can be updated anytime. To make it works, you have to make sure that the daemon that detect the metadata changed is running (the daemon is called the cfn-hup)
If you already use the Metadata and AWS::CloudFormation::Init, the data is not immediately being updated. As far I know, here is the condition the data to be change after change the Metadata value.
Reboot the instance
Run cfn-init command again with it's parameter
Waiting about 15 minutes, because the daemon to check the change in Metadata is checking the change once in 15 minutes.
I have a Lambda function invoked by S3 put events, which in turn needs to process the objects and write to a database on RDS. I want to test things out in my staging stack, which means I have a separate bucket, different database endpoint on RDS, and separate IAM roles.
I know how to configure the lambda function's event source and IAM stuff manually (in the Console), and I've read about lambda aliases and versions, but I don't see any support for providing operational parameters (like the name of the destination database) on a per-alias basis. So when I make a change to the function, right now it looks like I need a separate copy of the function for staging and production, and I would have to keep them in sync manually. All of the logic in the code would be the same, and while I get the source bucket and key as a parameter to the function when it's invoked, I don't currently have a way to pass in the destination stuff.
For the destination DB information, I could have a switch statement in the function body that checks the originating S3 bucket and makes a decision, but I hate making every function have to keep that mapping internally. That wouldn't work for the DB credentials or IAM policies, though.
I suppose I could automate all or most of this with the SDK. Has anyone set something like this up for a continuous integration-style deployment with Lambda, or is there a simpler way to do it that I've missed?
I found a workaround using Lambda function aliases. Given the context object, I can get the invoked_function_arn property, which has the alias (if any) at the end.
arn_string = context.invoked_function_arn
alias = arn_string.split(':')[-1]
Then I just use the alias as an index into a dict in my config.py module, and I'm good to go.
config[alias].host
config[alias].database
One thing I'm not crazy about is that I have to invoke my function from an alias every time, and now I can't use aliases for any other purpose without affecting this scheme. It would be nice to have explicit support for user parameters in the context object.