Capacity planner
Enter your traffic and latency. Get how many requests are in flight, and whether your database pool can take it.
How it works
- 1Little’s Law: requests in flight = requests per second × seconds per request.
- 2That one number sizes your connection pools, threads and memory. When latency goes up, it goes up with it.
- 3A database runs best with a small pool — roughly twice its CPU cores, across all your servers combined.
More demos
All demosQueryLite: a SQL database
A SQL database written from scratch — parser, B+ tree indexes, a query planner, joins and transactions — running live on 22,000 rows in your browser.
Raft consensus, live
The algorithm that keeps etcd and Kubernetes consistent. Five servers elect a leader and replicate a log — crash them and split the network while it runs.
Live collaborative editor
Real-time editing with no server. Three devices share a note — take one offline, edit everywhere, reconnect, and they merge. Open a second tab and it syncs live.