Simple Java program for running benchmarks on RDBMS.
Inspired by pg_bench, this software emulates a simple OLTP workload and measures TPS (transactions per second), average latency and standard deviation. Currently supported RDBMS are Oracle and PostgreSQL, but the Strategy design pattern used makes it easy to cover other engines.
$ java -jar JSqlBench.jar --help
Usage: JSqlBench [OPTIONS]
--engine=<engine> Database engine. Currently supported: Oracle and
Postgres
--host=<host> Database server's hostname (default: localhost)
--port=<port> Database server's port (default: 1521 for Oracle and
5432 for Postgres)
--dbname=<dbname> Database or instance name (SID)
--username=<username>
Username used to log in
--password=<password>
Password used to log in
--schema=<schema> Create objects in the specified namespace or schema,
rather than the default one
--tablespace=<tablespace>
Create objects in the specified tablespace, rather
than the default one
--nologging Create tables in nologging mode
--scale=<scale> Initialization scale factor, 1 = 100.000 rows. The
initialization scale factor should be at least as
large as the largest number of clients you intend
to test, else you'll mostly be measuring update
contention (default: 1)
--concurrency=<concurrency>
Number of concurrent clients simulated (default: 1)
--time=<time> Run the test for this many seconds. Never believe any
test that runs for only a few seconds, it is a good
practice to make the run last at least a few
minutes. In some cases you could need hours to get
numbers that are reproducible (default: 300)
--read-only Simulate a read only worlkoad
--help Print this help and exit