Devs@Home: The Quest for the Fastest Deployment Time - L Körbes
When a developer changes a line of code, how long does it take for that code to run in their development cluster? L Körbes, an expert in Kubernetes development tooling, has seen the answers to that question vary between 1 second and… infinity. So how do you get your team’s code to stay on the 1 second side? In this talk, L will explore every avenue to make our “from code change to process running” benchmark faster, and show you whether to…
…cache your deps with Docker layers or go mod,
remove debugging data to make Go binaries smaller,
pre-populate your container with the compiler cache,
keep a container running, sync code diffs to it, and recompile in place,
re-compile in place when you have a lot of memory and CPU,
re-compile when a lack of memory and CPU make the Go compiler sad,
use trickery to make binaries even smaller,
or, one last trick… how each of these differ when using a local cluster versus a remote one.
Expect an in-depth exploration with strategies and techniques you can put in practice to speed up your workflow for good!
https://entwickler.de/devs-at-home