Examples of Linux using FIO to test disk IO performance

Fio is a very good tool for testing iops to stress test and verify hardware, supporting 19 different I/O engines, including: Sync, mmap, Libaio, Posixaio, SG v3, splice, NULL, network, Syslet, Guasi, Solarisaio and so on.
FIO Official website: http://freecode.com/projects/fio/

First, installation

Can be installed directly yum, or download the compressed package to install

Yum Install Fio

Second, the order

1, Random read:

Fio-filename=/dev/sda1-direct=1-iodepth 1-thread-rw=randread-ioengine=psync-bs=4k-size=60g-numjobs=64-runtime=1 0-group_reporting-name=file

2, sequential reading:

Fio-filename=/dev/sda1-direct=1-iodepth 1-THREAD-RW=READ-IOENGINE=PSYNC-BS=4K-SIZE=60G-NUMJOBS=64-RUNTIME=10-GR Oup_reporting-name=file

3, Random write:

Fio-filename=/dev/sda1-direct=1-iodepth 1-thread-rw=randwrite-ioengine=psync-bs=4k-size=60g-numjobs=64-runtime= 10-group_reporting-name=file

4, sequential reading:

Fio-filename=/dev/sda1-direct=1-iodepth 1-thread-rw=write-ioengine=psync-bs=4k-size=60g-numjobs=64-runtime=10-g Roup_reporting-name=file

5. Mixed random reading and writing:

Fio-filename=/dev/sda1-direct=1-iodepth 1-thread-rw=randrw-rwmixread=30-ioengine=psync-bs=4k-size=60g-numjobs=6 4-runtime=10-group_reporting-name=file-ioscheduler=noop
Ps:
FILENAME=/DEV/SDA1 test file name, usually select the data directory of the disk you want to test
The Direct=1 test process bypasses the machine's own buffer. Make the test results more realistic
Rw=randwrite test for randomly written I/O
RW=RANDRW test Random Write and read I/O
Rw=randread test for Random read I/O
bs=4k Single IO block file size is 16k
bsrange=512-2048, the size range of the data block
size=60g This test file size is 60g, with every 4k IO test
NUMJOBS=64 the test thread for this time is 64
runtime=10 test time is 10 seconds, if not write 5g file is divided into 4k every time you finish
Ioengine=psync IO engine uses Pync method
Rwmixwrite=30 in mixed read-write mode, write accounted for 30%
Group_reporting about displaying the results, summarize the information for each process.
Furthermore
LOCKMEM=1G only use 1g memory for testing
Zero_buffers with 0 initialization system buffer
Nrfiles=8 the number of files generated per process.

Third, the example test

This example is a test for random reads, bw=1532.2kb/s in the text, iops=383 is the measured IOPS

[Root@localhost mapper]# fio-filename=/dev/sda1-direct=1-rw=randread-bs=4k-size 60g-numjobs=64-runtime=10-group_r Eporting-name=file

File: (g=0): Rw=randread, bs=4k-4k/4k-4k, Ioengine=sync, iodepth=1

...

File: (g=0): Rw=randread, bs=4k-4k/4k-4k, Ioengine=sync, iodepth=1

Fio 1.57

Starting processes

Jobs:4 (f=4): [______r_______________________r_r_______r_______________________] [3.5% done] [745k/0k/s] [181/0 IOPS] [Eta 12m:26s]s]

File: (Groupid=0, jobs=64): err= 0:pid=8717

READ:IO=15792KB, bw=1532.2kb/s, iops=383, runt= 10307msec

Clat (USEC): min=78, max=1303.3k, avg=99439.90, stdev=12973.11

Lat (usec): min=78, max=1303.3k, avg=99440.39, stdev=12973.10

BW (KB/S): min= 3, max=, per=2.52%, avg=38.61, stdev= 2.45

cpu:usr=0.01%, sys=0.68%, ctx=4026, majf=0, minf=1961

IO depths:1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%

submit:0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%

complete:0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%

Issued r/w/d:total=3948/0/0, short=0/0/0

Lat (usec): 100=0.13%, 250=0.61%, 500=0.25%, 750=0.08%, 1000=0.03%

Lat (msec): 2=0.03%, 4=0.10%, 10=0.76%, 20=1.85%, 50=10.61%

Lat (msec): 100=18.90%, 250=47.21%, 500=19.02%, 750=0.28%, 1000=0.10%

Lat (msec): 2000=0.05%

Run Status Group 0 (all jobs):

READ:IO=15792KB, aggrb=1532kb/s, minb=1568kb/s, maxb=1568kb/s, Mint=10307msec, maxt=10307msec

Disk stats (read/write):

SDA:IOS=3931/13, MERGE=0/13, ticks=584686/184, in_queue=584841, util=98.53%