We want a database choice anyone can check: the same open tests for every engine, every result public.
Building a database? Get it measured
Stroppy is an open project that measures databases the same way and publishes every result. We run the workloads, check protocol compatibility against the original engine, and link every number to the run behind it.
Compatibility
Engines claim the PostgreSQL or MySQL protocol; clients find the gaps in production, one query at a time.
Stroppy Conformity runs the same checks on the original and on the engine, and keeps what each answered: the request, the original’s reply, the engine’s reply.
Open the conformity mapSpeed
Each vendor picks its own workload, machine and settings, so two published numbers rarely measure the same thing.
We run the same TPC-C on one 32 vCPU · 128 GB node, every engine deployed and tuned the same way, and every row opens the report of its run.
See the ratingsClient
How a client sends a transaction changes the number it reports, often by more than the database does.
On the same OrioleDB, the same TPC-C ran at 4,410 transactions a second with 18 round trips each, and at 15,791 with one. The rating shows both.
How we found itEffort
Machines, deployment, tuning, monitoring and teardown come before the first number, and again for every engine.
Stroppy Cloud brings the stand up in your own cloud account, deploys and tunes the database, runs the test and keeps the report after the machines are gone.
Stroppy CloudEach engine against PostgreSQL itself, on the same checks. A score is how often the engine answers like the original; the line marks 80, where we call an engine compatible.
16 of 20PostgreSQL-compatible engines score under 80 of 100