Fio简介:
FIO是测试IOPS的非常好的工具,用来对硬件进行压力测试和验证,支持13种不同的I/O引擎,包括:sync,mmap, libaio, posixaio, SG v3, splice, null, network, syslet, guasi, solarisaio 等等。
使用源码安装 Fio
我们要去克隆 GitHub 上的仓库。安装所需的依赖,然后我们将会从源码构建应用。首先,确保我们安装了 Git 。
sudo apt-get install git
CentOS 用户可以执行下述命令:
sudo yum install git
现在,我们切换到 /opt
目录,并从 Github 上克隆仓库:
cd /opt git clone https://github.com/axboe/fio
你应该会看到下面这样的输出:
Cloning into 'fio'... remote: Counting objects: 24819, done. remote: Compressing objects: 100% (44/44), done. remote: Total 24819 (delta 39), reused 62 (delta 32), pack-reused 24743 Receiving objects: 100% (24819/24819), 16.07 MiB | 0 bytes/s, done. Resolving deltas: 100% (16251/16251), done. Checking connectivity... done.
现在,我们通过在 /opt
目录下输入下方的命令切换到 Fio 的代码目录:
cd fio
最后,我们可以使用下面的命令来使用 make
从源码构建软件:
# ./configure # make # make install
在 Ubuntu 上安装 Fio
对于 Ubuntu 和 Debian 来说, Fio 已经在主仓库内。你可以很容易的使用类似 yum
和 apt-get
的标准包管理器来安装 Fio。
对于 Ubuntu 和 Debian ,你只需要简单的执行下述命令:
sudo apt-get install fio
对于 CentOS/Redhat 你只需要简单执行下述命令。
在 CentOS ,你可能在你能安装 Fio 前需要去安装 EPEL 仓库到你的系统中。你可以通过执行下述命令来安装它:
sudo yum install epel-release -y
你可以执行下述命令来安装 Fio:
sudo yum install fio -y
使用 Fio 进行磁盘性能测试
现在 Fio 已经安装到了你的系统中。现在是时候看一些如何使用 Fio 的例子了。我们将进行随机写、读和读写测试。
执行随机写测试
执行下面的命令来开始。这个命令将要同一时间执行两个进程,写入共计 4GB( 4 个任务 x 512MB = 2GB) 文件:
sudo fio --name=randwrite --ioengine=libaio --iodepth=1 --rw=randwrite --bs=4k --direct=0 --size=512M --numjobs=2 --runtime=240 --group_reporting ... fio-2.2.10 Starting 2 processes randwrite: (groupid=0, jobs=2): err= 0: pid=7271: Sat Aug 5 13:28:44 2017 write: io=1024.0MB, bw=2485.5MB/s, iops=636271, runt= 412msec slat (usec): min=1, max=268, avg= 1.79, stdev= 1.01 clat (usec): min=0, max=13, avg= 0.20, stdev= 0.40 lat (usec): min=1, max=268, avg= 2.03, stdev= 1.01 clat percentiles (usec): | 1.00th=[ 0], 5.00th=[ 0], 10.00th=[ 0], 20.00th=[ 0], | 30.00th=[ 0], 40.00th=[ 0], 50.00th=[ 0], 60.00th=[ 0], | 70.00th=[ 0], 80.00th=[ 1], 90.00th=[ 1], 95.00th=[ 1], | 99.00th=[ 1], 99.50th=[ 1], 99.90th=[ 1], 99.95th=[ 1], | 99.99th=[ 1] lat (usec) : 2=99.99%, 4=0.01%, 10=0.01%, 20=0.01% cpu : usr=15.14%, sys=84.00%, ctx=8, majf=0, minf=26 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 : total=r=0/w=262144/d=0, short=r=0/w=0/d=0, drop=r=0/w=0/d=0 latency : target=0, window=0, percentile=100.00%, depth=1 Run status group 0 (all jobs): WRITE: io=1024.0MB, aggrb=2485.5MB/s, minb=2485.5MB/s, maxb=2485.5MB/s, mint=412msec, maxt=412msec Disk stats (read/write): sda: ios=0/0, merge=0/0, ticks=0/0, in_queue=0, util=0.00%
执行随机读测试
我们将要执行一个随机读测试,我们将会尝试读取一个随机的 2GB 文件。
sudo fio --name=randread --ioengine=libaio --iodepth=16 --rw=randread --bs=4k --direct=0 --size=512M --numjobs=4 --runtime=240 --group_reporting
你应该会看到下面这样的输出:
... fio-2.2.10 Starting 4 processes randread: Laying out IO file(s) (1 file(s) / 512MB) randread: Laying out IO file(s) (1 file(s) / 512MB) randread: Laying out IO file(s) (1 file(s) / 512MB) randread: Laying out IO file(s) (1 file(s) / 512MB) Jobs: 4 (f=4): [r(4)] [100.0% done] [71800KB/0KB/0KB /s] [17.1K/0/0 iops] [eta 00m:00s] randread: (groupid=0, jobs=4): err= 0: pid=7586: Sat Aug 5 13:30:52 2017 read : io=2048.0MB, bw=80719KB/s, iops=20179, runt= 25981msec slat (usec): min=72, max=10008, avg=195.79, stdev=94.72 clat (usec): min=2, max=28811, avg=2971.96, stdev=760.33 lat (usec): min=185, max=29080, avg=3167.96, stdev=798.91 clat percentiles (usec): | 1.00th=[ 2192], 5.00th=[ 2448], 10.00th=[ 2576], 20.00th=[ 2736], | 30.00th=[ 2800], 40.00th=[ 2832], 50.00th=[ 2928], 60.00th=[ 3024], | 70.00th=[ 3120], 80.00th=[ 3184], 90.00th=[ 3248], 95.00th=[ 3312], | 99.00th=[ 3536], 99.50th=[ 6304], 99.90th=[15168], 99.95th=[18816], | 99.99th=[22912] bw (KB /s): min=17360, max=25144, per=25.05%, avg=20216.90, stdev=1605.65 lat (usec) : 4=0.01%, 10=0.01%, 250=0.01%, 500=0.01%, 750=0.01% lat (usec) : 1000=0.01% lat (msec) : 2=0.01%, 4=99.27%, 10=0.44%, 20=0.24%, 50=0.04% cpu : usr=1.35%, sys=5.18%, ctx=524309, majf=0, minf=98 IO depths : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=100.0%, 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=0.1%, 32=0.0%, 64=0.0%, >=64=0.0% issued : total=r=524288/w=0/d=0, short=r=0/w=0/d=0, drop=r=0/w=0/d=0 latency : target=0, window=0, percentile=100.00%, depth=16 Run status group 0 (all jobs): READ: io=2048.0MB, aggrb=80718KB/s, minb=80718KB/s, maxb=80718KB/s, mint=25981msec, maxt=25981msec Disk stats (read/write): sda: ios=521587/871, merge=0/1142, ticks=96664/612, in_queue=97284, util=99.85%
最后,我们想要展示一个简单的随机读-写测试来看一看 Fio 返回的输出类型。
读写性能测试
下述命令将会测试 USB Pen 驱动器 (/dev/sdc1
) 的随机读写性能:
sudo fio --randrepeat=1 --ioengine=libaio --direct=1 --gtod_reduce=1 --name=test --filename=random_read_write.fio --bs=4k --iodepth=64 --size=4G --readwrite=randrw --rwmixread=75
下面的内容是我们从上面的命令得到的输出:
fio-2.2.10 Starting 1 process Jobs: 1 (f=1): [m(1)] [100.0% done] [217.8MB/74452KB/0KB /s] [55.8K/18.7K/0 iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=8475: Sat Aug 5 13:36:04 2017 read : io=3071.7MB, bw=219374KB/s, iops=54843, runt= 14338msec write: io=1024.4MB, bw=73156KB/s, iops=18289, runt= 14338msec cpu : usr=6.78%, sys=20.81%, ctx=1007218, majf=0, minf=9 IO depths : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.1%, 32=0.1%, >=64=100.0% 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=0.1%, >=64=0.0% issued : total=r=786347/w=262229/d=0, short=r=0/w=0/d=0, drop=r=0/w=0/d=0 latency : target=0, window=0, percentile=100.00%, depth=64 Run status group 0 (all jobs): READ: io=3071.7MB, aggrb=219374KB/s, minb=219374KB/s, maxb=219374KB/s, mint=14338msec, maxt=14338msec WRITE: io=1024.4MB, aggrb=73156KB/s, minb=73156KB/s, maxb=73156KB/s, mint=14338msec, maxt=14338msec Disk stats (read/write): sda: ios=774141/258944, merge=1463/899, ticks=748800/150316, in_queue=900720, util=99.35%
以上就是
为各位朋友分享的 相关内容。想要了解更多Linux相关知识记得关注公众号“良许Linux”,或扫描下方二维码进行关注,更多 等着你!