Hi all: Recently i used fio for testing PCIe ssd performance.when i turned these text files into graph found that those data are recorded for per-jobs,not per-seconds.so the graph generated by fio_generate_plot can not fully reflect max performance results.Any advice for solve this problem,or whether have some other methods with fio can logging latency,bw and iops per-seconds .thanks a lot. [global] bs=4k ioengine=libaio #ioscheduler=noop iodepth=64 direct=1 runtime=9000 numjobs=32 randrepeat=0 filename=/dev/md0 norandommap group_reporting log_avg_msec=1200 [rand-read_4k] stonewall rw=randread write_bw_log=randread_4k write_lat_log=randread_4k write_iops_log=randread_4k disable_lat=1 disable_slat=1 [rand-write_4k] stonewall rw=randwrite write_bw_log=randwrite_4k write_lat_log=randwrite_4k write_iops_log=randwrite_4k disable_lat=1 disable_slat=1 [rand-read_16k] stonewall bs=16k rw=randread write_bw_log=randread_16k write_lat_log=randread_16k write_iops_log=randread_16k disable_lat=1 disable_slat=1 [rand-write_16k] stonewall bs=16k rw=randwrite write_bw_log=randwrite_16k write_lat_log=randwrite_16k write_iops_log=randwrite_16k disable_lat=1 disable_slat=1 [seq-read_4k] stonewall rw=read write_bw_log=read_4k write_lat_log=read_4k write_iops_log=read_4k disable_lat=1 disable_slat=1 [seq-write_4k] stonewall rw=write write_bw_log=write_4k write_lat_log=write_4k write_iops_log=write_4k disable_lat=1 disable_slat=1 bo.zhou DBA -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
