Discussions related to GitHub Actions
Has anyone looked at getting the output from a terraform Cloud plan using github actions and posting as a comment , sorta like Atlantis does?
Looks like it can return outputs from the current terraform state. Might not be the same plan, but still useful for potentionally adding output to merged pull request perhaps.
And one more question, should I combined all my linting tasks into a single workflow to improve performance/runtime? They don’t show up in the checks individually then. Should I keep them seperate in small pieces instead so each linting check shows up individually in the pull request?
fwiw, we’re using the github superlinter
that kind of validates your approach of combining them
But we also have an example like this one: https://github.com/cloudposse/packages/pull/392
what Drop codefresh pipeline for building docker image why Use github action instead for easier open source adoption
if you are ever curious here’s a guy that works at Microsoft that was inspired by the super linter but wanted to enhance it. This has some enhancements as linters are “structured objects” and more. Maybe you’ll find it interesting if you find you want more from the linter https://github.com/JustinGrote/Super-Duper-Linter
Inspired by github/super-linter, but using Powershell - JustinGrote/Super-Duper-Linter
GitHub Actions: Manual triggers with workflow_dispatch
First time setting up github actions and I’m
ya, they are super convenient
what did you use it for?
Set up a full deploy pipeline (build, publish ECR, deploy to ECS)
(btw, did you already start you new gig?)
Not yet, in between… so picked up some contracting, because I don’t know how to take a break
how is the smoke where you are?
btw, did you say you moved?
The one thing I need to figure out (outside scope of this channel) - what do I use to do an ECS deploy now (if github is down or if I want to revert to an older tag)
Haha how to do “ecs deploy” is a frequent topic
The general consensus for folks in our community is that using Terraform to do the deploy pretty much sucks.
By in large, our customers never ask for any “DR” solution for CI/CD
My guess is your customer is too small to warrant an investment in DR for GitHub.
If GitHub is down, then revert to WikiOps
Yea that’s what I’m going to do but instead just create make targets for all that stuff.. so combination of wikiops + makefile to ease the burden
Yea, so I’m in Santa Clarita now … smoke has been awful last few days, it seems like it’s getting better
still like a 35m drive from you Erik - if you are still in the same area
Yep! Still in Pasadena. But we are honestly itching to get out of the state.
Looking at Texas or Florida.
I can vouch for Miami
I guess both are going to have hurricane issues - but it’s not as bad as the news makes it most of the time
Ya, that’s true - trading one catastrophe for the next.
Good article on a topic we’ve talked about here regarding github actions and many repos/org usage. With the pace of change though Im betting this might change in the next year with some better centralized controls. We’ll see
With the beta release of GitHub Actions now available, is it time to retire your in house CI server?