> -----Original Message----- > From: [email protected] [mailto:[email protected]] On > Behalf Of Nguyen Viet Dung > Sent: Friday, April 03, 2015 6:43 AM > To: [email protected] > Subject: Need support about value IOPS Which FIO measured > > Dear > I'm a IT. I use FIO test performance my Hard disk on system of my > company. I searched about IOPS of hard disk on internet. And see that: > Spin Speed Interface Average IOPS > 5,900 rpm SATA 47 > 7,200 rpm SATA 75 > 10,000 rpm SATA / SAS 150 > 15,000 rpm SAS 175 > > So, when i use FIO measure my hardisk (western 500GB > I showed my result: IOPS read=34271, IOPS write=1198. This my result so > big. > I can't understand value IOPS of Fio. > > I hope receiver your reply. Thanks you so much. > > > testrw: (g=0): rw=rw, bs=4K-4K/4K-4K, ioengine=libaio, iodepth=8 > fio 1.59 > Starting 1 process > testrw: Laying out IO file(s) (1 file(s) / 100MB)
Using a file means you're going through a filesystem and the page cache. Using such a tiny file ensures that all writes are just going to system memory, and reads are getting cache hits and coming from system memory. > testrw: (groupid=0, jobs=1): err= 0: pid=8565 > read : io=77180KB, bw=137087KB/s, iops=34271 , runt= 563msec > slat (usec): min=1 , max=386 , avg= 3.57, stdev=15.60 > clat (usec): min=4 , max=1453 , avg=155.63, stdev=126.97 > lat (usec): min=5 , max=1456 , avg=159.35, stdev=130.95 The latency numbers also show this isn't accessing the HDD; HDD seek times are in ms, not us. Use the block device (e.g., /dev/sda) and access the full capacity of the drive to see the real HDD results. Also, if the HDD has its own write cache enabled, you'll get distorted results on writes. -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
