🪄 So we built something better. A lot better.

We designed a fully automated, secure, and cost-optimized heap dump retrieval system that transforms this manual process into a seamless developer experience.

Here’s how it works:

1. 🔧 Developer triggers an AWS SSM Automation Document
   → Inputs: ECS task ID, cluster name, and container name

2. 🚀 Behind the scenes:

   • A temporary AWS Batch EC2 job is spun up

   • It runs jcmd inside the container to generate the heap dump

   • The dump is uploaded directly to S3

   • A JIFA webapp container is booted with the dump mounted

   • A Cloudflare Tunnel securely exposes the JIFA UI

   • SSO authentication protects access

   • And here's the magic: the link auto-expires if the developer is inactive

Why this matters:

• ✅ No more manual toil for infra folks

• 🔐 Secure and auditable access for developers

• ⚡ Fast, self-service debugging

• 💸 Cost-efficient — powered by ephemeral infrastructure

• 🧠 Designed with developer experience in mind

This has already saved us a ton of time and made our on-call life a lot less noisy.

If your team is still doing heap dumps manually, it might be time to rethink that. 🙂

✉️ Like what you're reading?
Subscribe to the newsletter to keep getting hands-on engineering stories, system design wins, and DevOps insights — all from the trenches.