Could you remove rate_process=poisson and try, with the same profile
without rate_poisson, i am getting a bit less than 100 IOPS expected.
# fio -name=rate
--filename=/dev/mapper/3600601602d003e0002cde5580011c8fb
--ioengine=libaio --direct=1 --fill_device=1 --group_reporting
--rw=randwrite --rate_iops=100 --iodepth=8 --norandommap
--bsrange=8k-16k --
bssplit=8k/90:16k/10
avg-cpu: %user %nice %system %iowait %steal %idle
0.25 0.00 0.38 1.13 0.00 98.24
Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s
avgrq-sz avgqu-sz await r_await w_await svctm %util
dm-18 0.00 0.00 0.00 91.00 0.00 800.00
17.58 0.38 4.20 0.00 4.20 3.96 36.00
--
Srinivasa R Chamarthy
On Thu, Apr 6, 2017 at 12:09 AM, GV Govindasamy
<gv.govindasamy@xxxxxxxxxxx> wrote:
>
> Hello All
>
> I would like to know what I observe is a bug or by design that I don't understand. I expect to generate 100 random read ops or 100 random write ops in the following examples with 90% 8k and 10% 16k. FIO also appears to target for 100 IOPS, but ended up doing 2x in actual workload.
>
> Thanks,
> +GV
>
> =============================================
> $ ./fio-2.19 --version (compiled with: ./configure --build-static; make on CentOS release 6.7/3.10.0-229)
> fio-2.19
>
> =============================================
> $ cat output/fio/bssplit_readtest.fio
> [global]
> ioengine=libaio
> direct=1
> time_based
> norandommap
> group_reporting
> disk_util=0
> continue_on_error=all
> rate_process=poisson
>
>
> [db-oltp-w]
> bssplit=8k/90:16k/10,,
> size=128G
> filename=/dev/sdg
> rw=randread
> iodepth=8
> rate_iops=100
>
> =========================================
> $ iostat -x 2 sdg
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.13 0.00 0.13 1.63 0.00 98.12
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 191.00 0.00 3392.00 0.00 17.76 0.09 0.45 0.45 0.00 0.41 7.75
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.25 0.00 0.25 1.75 0.00 97.74
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 193.00 0.00 3424.00 0.00 17.74 0.08 0.42 0.42 0.00 0.39 7.50
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.13 0.00 0.25 1.88 0.00 97.74
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 213.50 0.00 3672.00 0.00 17.20 0.10 0.45 0.45 0.00 0.40 8.45 =========================================
>
> $ sudo ./fio-2.19 --runtime 120 --eta-newline=30 output/fio/bssplit_readtest.fio
> db-oltp-w: (g=0): rw=randread, bs=(R) 8192B-16.0KiB, (W) 4096B-4096B, (T) 4096B-4096B, ioengine=libaio, iodepth=8
> fio-2.19
> Starting 1 process
> Jobs: 1 (f=1), 0-100 IOPS: [r(1)][26.4%][r=1632KiB/s,w=0KiB/s][r=194,w=0 IOPS][eta 01m:29s]
> Jobs: 1 (f=1), 0-100 IOPS: [r(1)][51.2%][r=1648KiB/s,w=0KiB/s][r=190,w=0 IOPS][eta 00m:59s]
> Jobs: 1 (f=1), 0-100 IOPS: [r(1)][76.0%][r=1760KiB/s,w=0KiB/s][r=193,w=0 IOPS][eta 00m:29s]
> Jobs: 1 (f=1), 0-100 IOPS: [r(1)][100.0%][r=1816KiB/s,w=0KiB/s][r=202,w=0 IOPS][eta 00m:00s]
> db-oltp-w: (groupid=0, jobs=1): err= 0: pid=1605: Wed Apr 5 15:53:00 2017
> read: IOPS=200, BW=1765KiB/s (1808kB/s)(207MiB/120013msec)
> slat (usec): min=3, max=117, avg=30.53, stdev=11.20
> clat (usec): min=120, max=8658, avg=413.81, stdev=216.67
> lat (usec): min=139, max=8697, avg=444.34, stdev=218.71
> clat percentiles (usec):
> | 1.00th=[ 175], 5.00th=[ 221], 10.00th=[ 251], 20.00th=[ 290],
> | 30.00th=[ 338], 40.00th=[ 382], 50.00th=[ 406], 60.00th=[ 434],
> | 70.00th=[ 462], 80.00th=[ 498], 90.00th=[ 556], 95.00th=[ 620],
> | 99.00th=[ 756], 99.50th=[ 932], 99.90th=[ 3632], 99.95th=[ 4704],
> | 99.99th=[ 6368]
> lat (usec) : 250=9.84%, 500=70.75%, 750=18.34%, 1000=0.60%
> lat (msec) : 2=0.26%, 4=0.12%, 10=0.08%
> cpu : usr=0.37%, sys=0.84%, ctx=44140, majf=0, minf=60
> IO depths : 1=91.3%, 2=8.6%, 4=0.1%, 8=0.1%, 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 rwt: total=24072,0,0, short=0,0,0, dropped=0,0,0
> errors : total=0, first_error=0/<Success>
> latency : target=0, window=0, percentile=100.00%, depth=8
>
> Run status group 0 (all jobs):
> READ: bw=1765KiB/s (1808kB/s), 1765KiB/s-1765KiB/s (1808kB/s-1808kB/s), io=207MiB (217MB), run=120013-120013msec
> ========================================
>
> $ cat output/fio/bssplit_writetest.fio
> [global]
> ioengine=libaio
> direct=1
> time_based
> norandommap
> group_reporting
> disk_util=0
> continue_on_error=all
> rate_process=poisson
>
> [db-oltp-w]
> bssplit=,8k/90:16k/10,
> size=128G
> filename=/dev/sdg
> rw=randwrite
> iodepth=8
> rate_iops=100
> =========================================
> $ iostat -x 2 sdg
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.13 0.00 0.25 4.89 0.00 94.74
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 0.00 195.00 0.00 3480.00 17.85 0.28 1.43 0.00 1.43 1.06 20.65
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.13 0.00 0.38 4.38 0.00 95.12
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 0.00 202.50 0.00 3632.00 17.94 0.24 1.18 0.00 1.18 0.92 18.60
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.00 0.00 0.25 3.76 0.00 95.98
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 0.00 179.00 0.00 3072.00 17.16 0.21 1.15 0.00 1.15 0.92 16.45
>
> avg-cpu: %user %nice %system %iowait %steal %idle
> 0.13 0.00 0.25 4.52 0.00 95.11
> Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await r_await w_await svctm %util
> sdg 0.00 0.00 0.00 199.50 0.00 3472.00 17.40 0.25 1.24 0.00 1.24 0.98 19.55
> ==========================================
>
> $ sudo ./fio-2.19 --runtime 120 --eta-newline=30 output/fio/bssplit_writetest.fio
> db-oltp-w: (g=0): rw=randwrite, bs=(R) 4096B-4096B, (W) 8192B-16.0KiB, (T) 4096B-4096B, ioengine=libaio, iodepth=8
> fio-2.19
> Starting 1 process
> Jobs: 1 (f=1), 0-100 IOPS: [w(1)][26.4%][r=0KiB/s,w=1632KiB/s][r=0,w=194 IOPS][eta 01m:29s]
> Jobs: 1 (f=1), 0-100 IOPS: [w(1)][51.2%][r=0KiB/s,w=1648KiB/s][r=0,w=190 IOPS][eta 00m:59s]
> Jobs: 1 (f=1), 0-100 IOPS: [w(1)][76.0%][r=0KiB/s,w=1760KiB/s][r=0,w=193 IOPS][eta 00m:29s]
> Jobs: 1 (f=1), 0-100 IOPS: [w(1)][100.0%][r=0KiB/s,w=1816KiB/s][r=0,w=202 IOPS][eta 00m:00s]
> db-oltp-w: (groupid=0, jobs=1): err= 0: pid=1616: Wed Apr 5 15:57:04 2017
> write: IOPS=200, BW=1765KiB/s (1807kB/s)(207MiB/120014msec)
> slat (usec): min=5, max=125, avg=31.91, stdev=12.20
> clat (usec): min=500, max=26437, avg=1170.96, stdev=849.81
> lat (usec): min=516, max=26448, avg=1202.87, stdev=849.55
> clat percentiles (usec):
> | 1.00th=[ 644], 5.00th=[ 732], 10.00th=[ 788], 20.00th=[ 868],
> | 30.00th=[ 932], 40.00th=[ 996], 50.00th=[ 1048], 60.00th=[ 1128],
> | 70.00th=[ 1192], 80.00th=[ 1288], 90.00th=[ 1448], 95.00th=[ 1608],
> | 99.00th=[ 4832], 99.50th=[ 5920], 99.90th=[12480], 99.95th=[13760],
> | 99.99th=[26496]
> lat (usec) : 750=6.15%, 1000=34.93%
> lat (msec) : 2=56.35%, 4=1.15%, 10=1.27%, 20=0.11%, 50=0.03%
> cpu : usr=0.39%, sys=0.79%, ctx=38562, majf=0, minf=34
> IO depths : 1=80.2%, 2=19.3%, 4=0.4%, 8=0.1%, 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 rwt: total=0,24072,0, short=0,0,0, dropped=0,0,0
> errors : total=0, first_error=0/<Success>
> latency : target=0, window=0, percentile=100.00%, depth=8
>
> Run status group 0 (all jobs):
> WRITE: bw=1765KiB/s (1807kB/s), 1765KiB/s-1765KiB/s (1807kB/s-1807kB/s), io=207MiB (217MB), run=120014-120014msec
> ==========================================--
> To unsubscribe from this list: send the line "unsubscribe fio" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
