In a monorepo, I'd like to run deploy after merging to main. Common enough. What I'm thinking now, is to run a deploy on everything, something like

./gradlew deploy

, and everything already deployed will be UP-TO-DATE, so a no-op. (I'm not sure how it would know if it's already deployed, probably via a hash of the binary?) Is this a good idea to model this in the build system?

Probably depends on the exact circumstances. 🤷‍♂️

I think the actual question here is, to have my deploy task no-op if already deployed, where that check is network call against some image registry good idea in gradle?

I'm not sure Gradle would be a good tool for this. sure you could

tasks.deployFoo {
    onlyIf { getLiveVersion("foo") != fooVersion }
}

but you wouldn't have a good way of seeing which deployments would actually happen before running them (

--dry-run

shows everything as

SKIPPED

regardless of

onlyIf

), which is something other tools made for this purpose like Terraform can do

thank you both but turns out I just reivented gitops 😀