
Survive memory crashes without upgrading your instance
THIS IS NOT CLICKBAIT, JUST ACTUAL STUFF.
Was running a self-hosted GitHub Actions runner on a t3.micro (1 GB RAM) to build Flutter and Node apps. Every single pipeline run crashed. CloudWatch showed the same pattern every time: CPU spike, memory spike, process dead.
Turns out you don't need to upgrade the instance to fix this. To find out what I did instead (and where this fix would actually be a bad idea), read the full breakdown here: https://builder.aws.com/content/3I19RRzHfurmLkbJQfBtA3lLQjx/survive-memory-crashes-without-upgrading-your-instance