You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Stateful resources and those likely to be managed independently of cirrus should not be created by the default set of builtins. Rather, init should create a basic set of cloudformation resource templates in the project's cloudformation/ directory, and leave it to the user to remove/modify those resources as necessary.
Examples of such resources include:
S3 data buckets
VPC and all its related components
Dynamodb database (maybe?)
For the S3 buckets, we should simply dump out the current built-in buckets configs to cloudformation/s3.yml and call it day.
Stateful resources and those likely to be managed independently of cirrus should not be created by the default set of builtins. Rather,
init
should create a basic set of cloudformation resource templates in the project'scloudformation/
directory, and leave it to the user to remove/modify those resources as necessary.Examples of such resources include:
For the S3 buckets, we should simply dump out the current built-in buckets configs to
cloudformation/s3.yml
and call it day.For the VPC, things get a bit more complicated. The minimum set of VPC-related resources to create a functional cirrus deployment is not small. That said, AWS does provide a sample template that provides all these resources we can use with minimal tweaks: https://github.com/awsdocs/aws-lambda-developer-guide/blob/main/templates/vpc-privatepublic.yaml.
In fact, the only changes I found I needed to make to that template to be compatible with cirrus were confined to the VPC endpoint configs:
To make use of these resources, I made the following changes to my cirrus.yml:
I ran a test through using batch and it appeared all relevant services were accessible and everything worked as intended.
The text was updated successfully, but these errors were encountered: