Skip to main content

Database benchmarking as a service

Stroppy Cloud

Pick a database and a workload. Stroppy Cloud brings up the machines, tunes the engine, runs the load and hands you a report.

The machines come up in your own cloud account and go away when the run ends; the report stays. Compare runs, put them on a schedule, publish them to the rating, or launch them from your CI.

  • Your cloud account: Yandex Cloud or AWS
  • Stroppy and Graphene CI underneath
  • Public reports by link

Databases

Each engine is deployed the same way and tuned for the machine it gets, so runs compare engines, not set-ups. Or point a test at a database you already run.
  • PostgreSQLand OrioleDB
  • MySQLand MariaDB
  • Picodatadistributed, PostgreSQL protocol
  • YDBdeployed for the run, or a managed one
  • CockroachDBdistributed, PostgreSQL protocol
  • Externalyour own database, by its address

A run

Connect a provider once. A run brings up its stand in your account, deploys and tunes the database, runs Stroppy while you watch, keeps the report and tears the stand down; it can keep the stand for a while to look into a failure.
  1. Standmachines in Yandex Cloud or AWS
  2. Databasedeployed and tuned for the machine
  3. LoadTPC-C, as transactions or with procedures; TPC-B, TPC-H, TPC-DS
  4. Reportthroughput, latency, errors, a chart per segment
  5. Teardownthe machines go, the report stays

Compare, schedule, share

One run answers one question. A suite runs variants of a test to compare them; a schedule runs a test on cron to catch regressions. A report opens by a public link, and runs sent to the public rating land here:

API and CI

Everything the console does goes through a public REST API with an OpenAPI contract. A personal token launches a test from your pipeline; webhooks report each step of the run.
Terminal
curl -X POST \
'https://cloud.stroppy.io/api/v1/t/{tenant}/tests/{id}:launch' \
-H "Authorization: Bearer $STROPPY_TOKEN"

# webhook events
# run.started run.stage run.finished
# run.failed run.cancelled
# suite.started suite.cell_finished suite.finished

The other tools