I have used S3 and S3-compatible storage for a long time now, and have used both s3cmd and the AWS CLI tooling to sync, either between buckets or with a local filesystem. It generally "just works".

Except that when it doesn't, it's really hard to debug.

Continuing my 2023 advent blogging, today is a tip on getting NTP to work on Amazon AWS EC2 instances.

I have some content that I store in S3-compatible object storage, and wanted to be able to (a) push to that storage, and (b) serve items from that storage.

Easey-peasey: use the Flysystem AWS S3 adapter, point it to my storage, and be done!

Except for one monkey wrench: I'm using OpenSwoole.

AWS CodeDeploy is a tool for automating application deployments to EC2 instances and clusters. It can pull application archives from either S3 or GitHub, and then allows you to specify how to install, configure, and run the application via a configuration specification and optionally hook scripts. When setup correctly, it can provide a powerful way to automate your deployments.

I started looking into it because I wanted to try out my site on PHP 7, and do a few new things with nginx that I wasn't doing before. Additionally, I've accidently forgotten to deploy a few times in the past year after writing a blog post, and I wanted to see if I solve that situation; I'd really enjoyed the "push-to-deploy" paradigm of OpenShift and EngineYard in the past, and wanted to see if I could recreate it.

Enrico first pointed me to the service, and I was later inspired by a slide deck by Ric Harvey. The process wasn't easy, due to a number of things that are not documented or not fully documented in the AWS CodeDeploy documentation, but in the end, I was able to accomplish exactly that: push-to-deploy. This post details what I found, some recommendations on how to create your deployments, and ways to avoid some of the pitfalls I fell into.