INSIGHTS

Engineering notes

Measured findings from delivery work. Specific technologies, specific problems, numbers where we have them.

DATA

Postgres query tuning in production

Reading EXPLAIN ANALYZE against a live table, and which of its numbers actually predict a slow query.

9 MIN
DATA

Choosing Mongo vs Postgres, honestly

The access patterns that genuinely favour a document store, and the far more common ones that do not.

12 MIN
API

Rate limiting an API gateway

Token bucket against sliding window at the edge, and what each does to a client retrying badly.

7 MIN
CLOUD

Docker builds that don't take ten minutes

Layer ordering, cache mounts and multi-stage builds — measured on a real Node service.

8 MIN

4 POSTS

Want this applied to your stack?

Most of these findings came out of client delivery. We can run the same passes on your system.

Request a quote