cc439ec2eb
Add the python script that we use to run NVMe driver performance test for different workloads. The scripts outputs a csv file that you can import to Excel and plot performance for different qd, rw mix & io size. The sh script performs the prework and then calls the py script. Change-Id: Id6fd0ff28221d8cf8b2c8a77e47288f9739341fc Signed-off-by: John Kariuki <John.K.Kariuki@intel.com> Reviewed-on: https://review.gerrithub.io/363315 Tested-by: SPDK Automated Test System <sys_sgsw@intel.com> Reviewed-by: Daniel Verkamp <daniel.verkamp@intel.com> Reviewed-by: Jim Harris <james.r.harris@intel.com> |
||
---|---|---|
.. | ||
fio_test.conf | ||
README | ||
run_fio_test.py | ||
run_fio_test.sh |
These scripts are used to perform benchmark testing with fio. The run_fio_test.py is the main script that runs the performance test and parses the test results. Users can populate test parameters for different fio workloads in the lists (q_depth, io_size, workload_type, mix, core_mask and run_time) at the top of the run_fio_test.py script. The run_fio_test.py puts the test results in a csv file named <hostname>_<num ssds>_perf_output.csv. The run_fio_test.sh script demonstrates how to invoke the run_fio_test.py script with the input parameters: path_to_fio_conf, path_to_ioengine and num_ssds. The run_fio_test.sh script will call the SPDK setup.sh script to unbind NVMe devices from the kernel driver and bind them to the uio driver. We include a sample fio configuration file that includes the parameters we use in our test environment. The run_fio_test.py will append the NVMe devices to the end of the configuration file. The number of NVMe devices used is specified using the num_ssds parameter. Usage: ./run_fio_test.sh