Idempotency keys, explained with one retry storm
A payment endpoint, a slow database and a client that retried every 200 ms. What idempotency keys fix and what they don't.
11 min read
~/seattle $ whoami
I write about databases, distributed systems and the bugs that live between them.
Recent writing
all 8 postsA payment endpoint, a slow database and a client that retried every 200 ms. What idempotency keys fix and what they don't.
11 min read
Width is time, height is depth, colour is usually noise. A short guide to finding the slow part on the first try.
7 min read
An unbounded queue doesn't remove load, it postpones the outage. How to choose a bound and what to do when it's full.
12 min read
Transaction pooling breaks a few things quietly. A checklist to run before you put PgBouncer in front of an app.
6 min read
Two incidents, one cause: machines that disagreed about the time. The fields we added to every log line afterwards.
10 min read
Forty lines, a context and a WaitGroup. A worker pool that drains its jobs on SIGTERM instead of dropping them.
8 min read
$ subscribe
Leave your email and the next post from Stack & Heap comes to you.