Programmatically expose the output of CDK list #8436
Comments
benwainwright
added
feature-request
|
@benwainwright - I believe we shoudl track the use case in #601 That's an interesting point about cdk list taking 4 seconds. Can you help me understand the layout and organization of your infrastructure a bit better? I think that's worth following up on to see if there are optimizations we can make. |
…8515) Add a test to verify stacks are serialized to the manifest file in a topological order, based on their dependencies. By adding this test we are committing to maintain the topological order of the stacks in the manifest file, allowing users to consume it without needing to sort the stacks. Note that all of the CDK toolkit commands such as `cdk list` and `cdk deploy` already assume this and do not sort stacks upon execution. > see [`cdk list`](https://github.com/aws/aws-cdk/blob/master/packages/aws-cdk/lib/cdk-toolkit.ts#L264) and [`cdk deploy`](https://github.com/aws/aws-cdk/blob/master/packages/aws-cdk/lib/cdk-toolkit.ts#L111) for reference I have initially added the test in `cx-api` module but after more consideration I think it is better suited in the `core` module since we want to verify this behavior in the context of a CDK application. I have left both tests in this PR for the sake of discussion in the event the reviewer thinks this actually fits better in `cx-api`. helps #8436 ---- *By submitting this pull request, I confirm that my contribution is made under the terms of the Apache-2.0 license*
|
Closing in favour of #601 |
|
mrgrain
changed the title
Currently, I can get a list of stacks that will be generated by my project in the order they should be deployed (this is important; see below) by running the command
cdk listIt would be useful if the output from this command could be exposed as an api in a way that doesn't require full synthesis of the templates under the hood.
Use Case
On my team in the BBC, we are currently using the CDK to generate templates for 17 CodeBuild projects associated with various different Github repositories, along with three other templates that the others all depend on. Typically in the BBC, we deploy CloudFormation stacks via an internal service that acts as an intermediary between the developer or CI service and the AWS API.
When we first started developing the project that generates the CodeBuild templates, none of them had any dependencies on each other, and so a deployment solution was written that first runs
cdk synthto generate the templates, then deploys the templates in filesystem order.As our use case became more complicated, we started introducing dependencies between templates and we learned that, under the hood, this resulted in imports/exports between templates being implicitly generated
This obviously left us with the problem that we had no guarantee that we would deploy the templates in the correct order, and we quickly ran into deployment failures. We reached out to our AWS account team about this and after a quick call, we got a response identifying that we should use the output of
cdk list, since this not only lists the stacks to be generated, but does so in the order they should be deployed.While this solves our immediate problem (and so this is not a high priority issue for us), it would be really useful to be able to do this programmatically without having to shell out. This would make our tests much cleaner/simpler, and it is also worth noting that
cdk listtook approx 4 seconds when I ran it which seems surprisingly slow given what it does.This is a 🚀 Feature Request
The text was updated successfully, but these errors were encountered: