Closed ciaransweet closed 3 months ago
Given there's no change for downstream services now, with the re-rename, is this good to merge @anayeaye @botanical ?
Looks like this can go in @anayeaye @botanical - Ran OK in the job https://github.com/NASA-IMPACT/veda-deploy/actions/runs/10280359453
Towards #402
What this PR is
This PR ensures that outputs from CDK for the following services:
stac-api
,ingest-api
,raster-api
andstac-browser
are available and deterministically named.Prior to this change, if you piped the outputs to a file, you get a combination of
export_name
and a hash, which will cause issues when trying to retrieve values from the outputs for later jobs.To get around this, I've specified the
key
parameter which explicitly tells CDK to save the outputs under a name without any logical ID logic from CDK.I've also renamed the outputs file for this action to be
veda-backend-cdk-outputs.json
just so it's clear which stack it's for and prevents any clashes from further stacks we deploy.How you can test it
You can run a
cdk deploy/diff
locally and notice that the outputs are generated with the deterministic names, you can see the difference when compared to outputs I've not changed yet:You can see that the arns are still logical output id's and so is the secret name.