HGST HUSMM1640 Typical Performance?

Propaganda

Member
Dec 6, 2017
54
18
8
41
I am not sure what I was expecting but the 4k performance on these drives appears poor? Or perhaps it is my HP h240 HBA? Even worse in ATTO than my 840 pro running the OS.
1617053803294.png
1617054408675.png
1617054784198.png
1617054969187.png
 

Propaganda

Member
Dec 6, 2017
54
18
8
41
More data, I ran the drive in 2 different machines under freenas. One machine with a HP H240 HBA (SAS3) the other with a LSI 9217-4I4E HBA (SAS2). This is looking like an HBA issue?

HP H240 HBA:
Code:
root@truenas[/mnt/t1]# fio --filename=test --sync=1 --rw=randread --bs=4k --numjobs=1 --iodepth=4 --group_reporting --name=test --filesize=10G --runtime=300
test: (g=0): rw=randread, bs=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B,ioengine=psync, iodepth=4
fio-3.19
Starting 1 process
test: Laying out IO file (1 file / 10240MiB)
Jobs: 1 (f=1): [r(1)][100.0%][r=282MiB/s][r=72.2k IOPS][eta 00m:00s]
test: (groupid=0, jobs=1): err= 0: pid=3438: Mon Mar 29 18:11:32 2021
read: IOPS=44.0k, BW=176MiB/s (184MB/s)(10.0GiB/58295msec)
clat (nsec): min=1970, max=1190.1k, avg=21175.50, stdev=5201.83
lat (usec): min=2, max=1190, avg=21.27, stdev= 5.20
clat percentiles (nsec):
|  1.00th=[ 2832],  5.00th=[ 3504], 10.00th=[21120], 20.00th=[21632],
| 30.00th=[21888], 40.00th=[21888], 50.00th=[22144], 60.00th=[22400],
| 70.00th=[22656], 80.00th=[22912], 90.00th=[23168], 95.00th=[23680],
| 99.00th=[34560], 99.50th=[35584], 99.90th=[36608], 99.95th=[37120],
| 99.99th=[38656]
bw (  KiB/s): min=169245, max=280303, per=98.96%, avg=178003.11, stdev=12452.94, samples=109
iops        : min=42311, max=70075, avg=44500.43, stdev=3113.20, samples=109
lat (usec)   : 2=0.01%, 4=6.72%, 10=0.06%, 20=0.05%, 50=93.17%
lat (usec)   : 100=0.01%, 250=0.01%, 750=0.01%, 1000=0.01%
lat (msec)   : 2=0.01%
cpu          : usr=6.44%, sys=93.55%, ctx=1046, majf=0, minf=2
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=2621440,0,0,0 short=0,0,0,0 dropped=0,0,0,0
latency   : target=0, window=0, percentile=100.00%, depth=4
Run status group 0 (all jobs):
   READ: bw=176MiB/s (184MB/s), 176MiB/s-176MiB/s (184MB/s-184MB/s), io=10.0GiB(10.7GB), run=58295-58295msec
LSI 9217-4I4E HBA:
Code:
root@freenas0[/mnt/t0]# fio --filename=test --sync=1 --rw=randread --bs=4k --numjobs=1 --iodepth=4 --group_reporting --name=test --filesize=10G --runtime=300
test: (g=0): rw=randread, bs=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B,ioengine=psync, iodepth=4
fio-3.16
Starting 1 process
test: Laying out IO file (1 file / 10240MiB)
Jobs: 1 (f=1): [r(1)][91.7%][r=660MiB/s][r=169k IOPS][eta 00m:02s]
test: (groupid=0, jobs=1): err= 0: pid=58047: Mon Mar 29 20:35:40 2021
read: IOPS=119k, BW=465MiB/s (488MB/s)(10.0GiB/22001msec)
clat (usec): min=2, max=132, avg= 7.70, stdev= 8.25
lat (usec): min=2, max=133, avg= 7.77, stdev= 8.25
clat percentiles (nsec):
|  1.00th=[ 2736],  5.00th=[ 2992], 10.00th=[ 3056], 20.00th=[ 3120],
| 30.00th=[ 3184], 40.00th=[ 3248], 50.00th=[ 3344], 60.00th=[ 3536],
| 70.00th=[ 3696], 80.00th=[22144], 90.00th=[22912], 95.00th=[23424],
| 99.00th=[24448], 99.50th=[24960], 99.90th=[36096], 99.95th=[36608],
| 99.99th=[37632]
| 99.00th=[24448], 99.50th=[24960], 99.90th=[36096], 99.95th=[36608],
| 99.99th=[37632]
bw (  KiB/s): min=416644, max=816538, per=95.28%, avg=454120.14, stdev=78569.59, samples=43
iops        : min=104161, max=204134, avg=113529.77, stdev=19642.33, samples=43
lat (usec)   : 4=76.45%, 10=1.30%, 20=0.09%, 50=22.15%, 100=0.01%
lat (usec)   : 250=0.01%
cpu          : usr=11.13%, sys=88.86%, ctx=427, majf=0, minf=4
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=2621440,0,0,0 short=0,0,0,0 dropped=0,0,0,0
latency   : target=0, window=0, percentile=100.00%, depth=4
Run status group 0 (all jobs):
   READ: bw=465MiB/s (488MB/s), 465MiB/s-465MiB/s (488MB/s-488MB/s), io=10.0GiB(10.7GB), run=22001-22001msec
 

Propaganda

Member
Dec 6, 2017
54
18
8
41
Just got in new controller (LSI 9340-8i) and cables put in the same machine as the HP H240. Much better results, the HP H240 is anemic in comparison:
1617666325159.png
 
  • Like
Reactions: hitmanbabyvn