LinkedIn Series Announcement: The Case for Local Storage
Storage Series — Introduction
I just published a four-part series on storage. Not because it is a glamorous topic. Because it is costing you money every month and slowing your users down every day.
Most engineering teams accept EBS as the default. It is there, it works, the bill arrives and gets paid. Nobody questions it because questioning it means admitting there might be a better answer that requires actual work to implement.
There is a better answer.
Over the last four weeks I walked through the complete case for moving storage back onto the machine — from the user experience problem that nobody connects to infrastructure, through the architecture that solves it, to the honest decision framework that tells you when it makes sense and when it does not.
Here is what the series covers:
Article 1 — You’re Paying for the Wire
EBS is a network block device. Every IO your application makes travels across a network and comes back. That round trip is the spinner your users are watching. Local NVMe eliminates it entirely.
Article 2 — Four Drives. Production Grade. Nothing Extra to Pay.
ZFS RAID-10 on ECC hardware. What it gives you that EBS cannot — ARC caching, transparent compression, instant snapshots, end-to-end checksumming — and why ECC RAM is the foundation everything else depends on.
Article 3 — Same Database. Same Queries. Different Universe.
PostgreSQL on EBS gp3 versus PostgreSQL on local NVMe RAID-10. The latency numbers, the throughput gap, and why PostgreSQL specifically rewards fast local storage more than almost any other database.
Article 4 — When to Cut the Wire. And When Not To.
The honest part. When bare-metal local storage is the clear engineering answer, and when cloud storage is still the right call. No tribalism, just the decision framework your cloud vendor will not give you.
The hardware pays for itself in six months. The performance gap is an order of magnitude. The web got slower when storage moved off the machine.
Moving it back is a decision, not a revolution.
Links to all four articles in the comments below.