Postgres Until It Hurts: The Operational Case for One Database Doing Several Jobs
The pendulum swung back to one relational database for documents, queues, search and vectors: the case for it, and where a specialised system still wins.
6 min read
5 posts, newest first.
The pendulum swung back to one relational database for documents, queues, search and vectors: the case for it, and where a specialised system still wins.
6 min read
Caching at the edge moves bytes closer to users. Running code at the edge moves the code away from the data. What that trade costs, and when it genuinely pays.
5 min read
Deploys taught us every change is reversible. Schema and data migrations are not, and teams discover which step was one-way during the incident it caused.
6 min read
Most teams reaching for a distributed log want a queue, or a table with a status column. What a log costs to run, and the four properties that justify one.
5 min read
Round-robin, least-connections, IP hash, weighted and random: what each does to your tail latency, and when the popular default is the wrong choice.
5 min read