Collection of Cloud Formation custom resources written in python 3.6, result of months of continuous efforts to automate infrastructure management trough AWS CloudFormation. You may find some of these CloudFormation resources obsolete, as AWS team fills in the gaps. There is also some more complex ones, or developed to suite specific needs, such as copying s3 objects between buckets
Collection of custom resources for Node.js can be found here
Clone the repository, install the dependencies, and zip up the folder as Lambda Deployment package.
Note that boto3
is not packaged as dependency, as it is available in Lambda runtime itself by default.
git clone https://github.com/base2Services/cloudformation-custom-resources-python.git && \
cd cloudformation-custom-resources-python && \
docker run --rm -v $PWD:/src -w /src python:3.6 pip install -r requirements.txt -t . && \
zip -r ~/cfn-ccr-python3.6.zip .
It is easy to create sub-stacks in CloudFormation as long as they are in same region.
In some cases, there is need to create stack in region different than region where
parent stack is being create, or for example, to create same stack in multiple regions.
Such (sub)stack lifecycle can be controlled via custom resource having it's code in
regional-cfn-stack
folder
handler: regional-cfn-stack/handler.lambda_handler
runtime: python3.6
Required parameters:
Region
- AWS Region to create stack inStackName
- Name of the stack to be createdTemplateUrl
- S3 Url of stack templateCapabilities
- Comma seperated list of capabilities. Set to empty value if no IAM capabilities required.EnabledRegions
- Comma separated list of regions that stack is allowed to be created in. Useful when passing this list is template parameters.
Optional parameters:
StackParam_Key
- Will pass value of this param down to stack'sKey
parameterOnFailure
- Behaviour on stack creation failure. Accepted values areDO_NOTHING
,ROLLBACK
andDELETE
This custom resource allows copying from source to destination s3 buckets. For source, if you provide prefix
(without trailing slash), all objects under that prefix will be copied. Alternatively, if you provide s3 object
with *.zip
extensions, this object will be unpacked before it's files are unpacked to target bucket / prefix.
Please note that this lambda function design does not include recursive calls if lambda is timing out, thus it does not
permit mass file unpacking, but is rather designed for deployment of smaller files, such as client side web applications.
handler: 3-copy/handler.lambda_handler
runtime: python3.6
Required parameters:
Source
- Source object/prefix/zip-file ins3://bucket-name/path/to/prefix/or/object.zip
formatDestination
- Destination bucket and prefix ins3://bucket-name/destination-prefix
formatCannedAcl
- Canned ACL for created objects in destination No optional parameters.
This custom resource allows create/update/delete match regex rules with regex a pattern set.
handler: waf_regex/handler.lambda_handler
runtime: python3.6
Required parameters:
Name
- Custom Resource name.RegexPatterns
- List format, regex pattern to match.Type
- The part of the web request that you want AWS WAF to search for a specified stringData
- Data such as when the value of Type is HEADER , enter the name of the header that you want AWS WAF to search, for example, User-Agent or RefererTransform
- Text transformations eliminate some of the unusual formatting that attackers use in web requests in an effort to bypass AWS WAF. Implementation require to be serialised with other waf condition.
This custom resource creates a AmazonMQ broker instance.
NOTE: This resource cannot be updated. If a change to the instance is required such as Instance Type, a new broker resource must be created.
handler: amazon-mq-broker/handler.lambda_handler
runtime: python3.6
Required parameters:
Name
- Unique name given to the brokerSecurityGroups
- Array of security group idsSubnets
- Array of subnets idsMultiAZ
- String boolean [ 'true', 'false' ]InstanceType
- valid values [ 'mq.t2.micro', 'mq.m4.large' ]Username
- Username for the amq userPassword
- Password for the amq user. Must be 12-250 characters long
No optional parameters.
Returned Values:
Active
- Active AmazonMQ endpointStanby
- Standby AmazonMQ endpointBrokerId
- Id of the AmazonMQ BrokerBrokerArn
- Arn of the broker
IAM Permissions:
{
"Statement":
[
{
"Effect": "Allow",
"Action":
[
"mq:*",
"ec2:CreateNetworkInterface",
"ec2:CreateNetworkInterfacePermission",
"ec2:DeleteNetworkInterface",
"ec2:DeleteNetworkInterfacePermission",
"ec2:DetachNetworkInterface",
"ec2:DescribeInternetGateways",
"ec2:DescribeNetworkInterfaces",
"ec2:DescribeNetworkInterfacePermissions",
"ec2:DescribeRouteTables",
"ec2:DescribeSecurityGroups",
"ec2:DescribeSubnets",
"ec2:DescribeVpcs",
"logs:CreateLogGroup",
"logs:CreateLogStream",
"logs:PutLogEvents",
"lambda:InvokeFunction"
],
"Resource": ["*"]
}
]
}
This custom resource generates a random string [a-z][A-Z][0-9]
a definable length. The string is then return to the cfn stack and can then be passed into other resources requiring a password. The resource can be updated generating a new password and updating the SSM parameter and returning the new password by passing a dummy parameter into the custom resource.
handler: ssm-secure-parameter/handler.lambda_handler
runtime: python3.6
Required parameters:
Path
- SSM parameter path e.g./app/env/password
Optional parameters:
Length
- Length of the auto generated password. Defaults to 16 characters
Returned Values:
Password
- The password generated by the resource
IAM Permissions:
{
"Statement":
[
{
"Effect": "Allow",
"Action":
[
"ssm:PutParameter",
"ssm:DeleteParameter",
"logs:CreateLogGroup",
"logs:CreateLogStream",
"logs:PutLogEvents"
],
"Resource": ["*"]
}
]
}