Unexpected big latency during FIO testing

Open ningqingqing opened this issue 3 years ago • 0 comments

Please acknowledge the following before creating a ticket

[x] I have read the GitHub issues section of REPORTING-BUGS.

Description of the bug: <We have run random write test and record latency numbers, but we found some unexpected big latency during test.>

Environment: <Linux, CentOS 7.7.1908>

fio version: <FIO-3.20>

Reproduction steps <We parsed fio hist log using script fiologparser_hist.py, and the log shows there are 88 io exceed 1ms, but actually only one of them was detected with our internal latency monitor tool. Here are parameters we use: [global] max_latency=8s percentile_list=1:5:10:20:30:40:50:60:70:80:90:95:99:99.9:99.99:99.999:99.9999:99.99999:99.999999

[wrkldrnd_write_4k_QD1] readwrite=randwrite numjobs=1 iodepth=1 ioengine=libaio bs=4k ba=4k randrepeat=0 norandommap refill_buffers=0 direct=1 size=100% time_based ramp_time=330 runtime=10800 group_reporting=1 log_avg_msec=1000 log_hist_msec=1000 disk_util=1

Aug 08 '22 06:08 ningqingqing