Scroll to navigation

SYSBENCH(1) Benchmark tool for database systems SYSBENCH(1)


sysbench - multi-threaded benchmark tool for database systems


sysbench [common-options] --test=name [test-options] <command>
Commands: prepare run cleanup help version


SysBench is a modular, cross-platform and multi-threaded benchmark tool for evaluating OS parameters that are important for a system running a database under intensive load.

The idea of this benchmark suite is to quickly get an impression about system performance without setting up complex database benchmarks or even without installing a database at all.

The design is very simple. SysBench runs a specified number of threads and they all execute requests in parallel. The actual workload produced by requests depends on the specified test mode. You can limit either the total number of requests or the total time for the benchmark, or both.

Compiled-in database drivers:

  • mysql - MySQL driver
  • pgsql - PostgreSQL driver

Compiled-in tests:

  • fileio - File I/O test
  • cpu - CPU performance test
  • memory - Memory functions speed test
  • threads - Threads subsystem performance test
  • mutex - Mutex performance test

Available test modes are implemented by compiled-in modules, and SysBench was designed to make adding new test modes an easy task. Each test mode may have additional (or workload-specific) options.


number of threads to use [1]
limit for total number of events [0]
limit for total execution time in seconds [10]
number of seconds to wait after the --time limit before forcing shutdown, or 'off' to disable [off]
size of stack per thread [64K]
average transactions rate. 0 for unlimited rate [0]
periodically report intermediate statistics with a specified interval in seconds. 0 disables intermediate reports [0] --report-checkpoints=[LIST,...] dump full statistics and reset all counters at specified points in time. The argument is a list of comma-separated values representing the amount of time in seconds elapsed from start of test when report checkpoint(s) must be performed. Report checkpoints are off by default. []
print more debugging info [off]
perform validation checks where possible [off]
print help and exit [off]
print version and exit [off]
File containing command line options
deprecated alias for --rate [0]
deprecated alias for --events [0]
deprecated alias for --time [0]
deprecated alias for --threads [1]


random numbers distribution {uniform,gaussian,pareto} [special]
seed for random number generator. When 0, the current time is used as a RNG seed. [0]
parameter h for pareto distribution [0.2]
shape parameter (exponent, theta) for the Zipfian distribution [0.8]


verbosity level {5 - debug, 0 - only critical messages} [3]
percentile to calculate in latency statistics (1-100). Use the special value of 0 to disable percentile calculations [95]
print latency histogram in report [off]


specifies database driver to use ('help' to get list of available drivers) [mysql]
prepared statements usage mode {auto, disable} [auto]
print database-specific debug information [off]


MySQL server host [localhost]
MySQL server port [3306]
MySQL socket
MySQL user [sbtest]
MySQL password []
MySQL database name [sbtest]
use SSL connections, if available in the client library [off]
use specific cipher for SSL connections []
use compression, if available in the client library [off]
trace all client library calls [off] --mysql-ignore-errors=[LIST,...] list of errors to ignore, or "all" [1213,1020,1205]
Dry run, pretend that all MySQL client API calls are successful without executing them [off]


PostgreSQL server host [localhost]
PostgreSQL server port [5432]
PostgreSQL user [sbtest] --pgsql-password=STRING PostgreSQL password []
PostgreSQL database name [sbtest]


number of files to create [128]
block size to use in all IO operations [16384]
total size of files to create [2G]
test mode {seqwr, seqrewr, seqrd, rndrd, rndwr, rndrw}
file operations mode {sync,async,mmap} [sync]
number of asynchronous operatons to queue per thread [128] --file-extra-flags=[LIST,...] list of additional flags to use to open files {sync,dsync,direct} []
do fsync() after this number of requests (0 - don't use fsync()) [100]
do fsync() after each write operation [off]
do fsync() at the end of test [on]
which method to use for synchronization {fsync, fdatasync} [fsync]
merge at most this number of IO requests if possible (0 - don't merge) [0]
reads/writes ratio for combined test [1.5]


--cpu-max-prime=N upper limit for primes generator [10000]


size of memory block for test [1K]
total size of data to transfer [100G]
memory access scope {global,local} [global]
allocate memory from HugeTLB pool [off]
type of memory operations {read, write, none} [write] --memory-access-mode=STRING memory access mode {seq,rnd} [seq]


--thread-yields=N number of yields to do per request [1000]
number of locks per thread [8]


total size of mutex array [4096] --mutex-locks=N number of mutex locks to do per thread [50000] --mutex-loops=N number of empty loops to do outside mutex lock [10000]


Upstream provided html-formatted manual, describing the features of sysbench: /usr/share/doc/sysbench/manual.html
05 June 2020