linux stress, impose work on specific disks

Question

Have a server running centos 7.6, and it has 4 ssd's as Raid-0 mounted as /scratch/

I have the linux program stress-1.0.4-16 and I just learned stress-ng existed.

Is there a way with stress to tell it to do I/O operations to stress a specific set of disks such as my 4 disk Raid-0? Or does it only work on whatever the root file system is such as /tmp? And if that's the case I've done systemctl enable tmp.mount which means the /tmp folder is no longer on disk negating the disk i/o function of stress ?

Answer 1

You just need to cd to a directory in the disk drive you want to stress. Using -v option confirms it creates the stress files on the current directory.

# stress -v -d 1 
stress: info: [29736] dispatching hogs: 0 cpu, 0 io, 0 vm, 1 hdd
stress: dbug: [29736] using backoff sleep of 3000us
stress: dbug: [29736] --> hoghdd worker 1 [29737] forked
stress: dbug: [29737] seeding 1048575 byte buffer with random data
stress: dbug: [29737] opened ./stress.yPWMGk for writing 1073741824 bytes
stress: dbug: [29737] unlinking ./stress.yPWMGk
stress: dbug: [29737] fast writing to ./stress.yPWMGk

I also suggest you use fio instead if you need a more accurate tool.

Answer 2

Stress test using fio:

RedHat/CentOS:

sudo yum -y install fio

Debian/Ubuntu:

sudo apt install fio

a 10 min benchmark:

$ fio --name=test --ioengine=sync --rw=randwrite --bs=4k --numjobs=1 --size=1G --runtime=10m --time_base

test: (g=0): rw=randwrite, bs=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B, ioengine=sync, iodepth=1
fio-3.12
Starting 1 process
test: Laying out IO file (1 file / 1024MiB)
Jobs: 1 (f=1): [w(1)][100.0%][eta 00m:00s]                           
test: (groupid=0, jobs=1): err= 0: pid=924807: Wed Aug  6 11:04:48 2025
  write: IOPS=44.3k, BW=173MiB/s (181MB/s)(102GiB/603625msec); 0 zone resets
    clat (nsec): min=1742, max=1431.1k, avg=3418.70, stdev=1889.93
     lat (nsec): min=1815, max=1431.2k, avg=3530.50, stdev=1937.75
    clat percentiles (nsec):
     |  1.00th=[ 2192],  5.00th=[ 2768], 10.00th=[ 2832], 20.00th=[ 2896],
     | 30.00th=[ 2960], 40.00th=[ 2992], 50.00th=[ 3024], 60.00th=[ 3088],
     | 70.00th=[ 3152], 80.00th=[ 3344], 90.00th=[ 4512], 95.00th=[ 5536],
     | 99.00th=[ 9408], 99.50th=[10048], 99.90th=[13376], 99.95th=[14400],
     | 99.99th=[19072]
   bw (  KiB/s): min=13536, max=1168848, per=100.00%, avg=623553.31, stdev=302011.17, samples=343
   iops        : min= 3384, max=292212, avg=155888.34, stdev=75502.79, samples=343
  lat (usec)   : 2=0.10%, 4=83.58%, 10=15.79%, 20=0.52%, 50=0.01%
  lat (usec)   : 100=0.01%, 250=0.01%, 500=0.01%, 750=0.01%, 1000=0.01%
  lat (msec)   : 2=0.01%
  cpu          : usr=5.27%, sys=21.47%, ctx=39684, majf=0, minf=203
  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 rwts: total=0,26738689,0,0 short=0,0,0,0 dropped=0,0,0,0
     latency   : target=0, window=0, percentile=100.00%, depth=1

Run status group 0 (all jobs):
  WRITE: bw=173MiB/s (181MB/s), 173MiB/s-173MiB/s (181MB/s-181MB/s), io=102GiB (110GB), run=603625-603625msec

Disk stats (read/write):
    md2: ios=0/527583, merge=0/0, ticks=0/0, in_queue=0, util=0.00%, aggrios=0/533334, aggrmerge=0/5732, aggrticks=0/2386555, aggrin_queue=2356744, aggrutil=69.81%
  sdb: ios=0/533561, merge=0/5796, ticks=0/1995883, in_queue=1971388, util=53.01%
  sda: ios=0/533108, merge=0/5669, ticks=0/2777227, in_queue=2742100, util=69.81

In this case fio writes random data into file test.0.0 located in current work dir. The filesystem uses 2 SSDs in RAID1 (md2). Ideally both drives should have the same performance characteristic.