lkml.org 
[lkml]   [2020]   [Apr]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [LKP] [btrfs] 8d47a0d8f7: fio.write_bw_MBps -28.6% regression
From
Date
This regression happened on "lkp-hsw-ep2", but “lkp-hsw-ep2” has been 
removed from the LKP tbox, so we retest the regression on another tbox
“lkp-hsw-ep4”, their hardware is almost the same, "72 threads Intel(R)
Xeon(R) CPU E5-2699 v3 @ 2.30GHz with 256G memory". The test result is
in the attached file compare.txt. The regression is -6.7% for v5.6, the
origin regression is -8.9%, it is not serious as it test on
"lkp-hsw-ep2" before.


On 4/10/2020 6:14 PM, David Sterba wrote:
> On Fri, Apr 10, 2020 at 02:44:55PM +0800, Qu Wenruo wrote:
>> On 2020/4/10 下午2:34, Xing Zhengjun wrote:
>>> Hi Wenruo,
>>>
>>>    We test it in v5.6, the issue still exist, do you have time to take a
>>> look at this? Thanks.
>>
>> This is expected.
>>
>> The extra check brings new overhead mostly equal to another CRC32 run.
>>
>> We believe it's worthy, as our read time tree checker has exposed quite
>> some bit flip corruption.
>
> The test probably runs on a PMEM device so there's no slowdown from the
> actual IO and the in-memory checks are measurable, though 28% is a lot,
> I'd expect something like 5-10% at most.
>

--
Zhengjun Xing
=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/debug-setup/disk/fs/runtime/nr_task/time_based/rw/bs/ioengine/test_size/cpufreq_governor/ucode:
lkp-hsw-ep4/fio-basic/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/test/2pmem/btrfs/200s/50%/tb/randwrite/4k/libaio/100G/performance/0x43

commit:
ff2ac107fae2440b6877c615c0ac788d2a106ed7
8d47a0d8f7947422dd359ac8e462687f81a7a137
v5.6

ff2ac107fae2440b 8d47a0d8f7947422dd359ac8e46 v5.6
---------------- --------------------------- ---------------------------
%stddev %change %stddev %change %stddev
\ | \ | \
0.44 ± 7% -0.1 0.32 ± 4% -0.1 0.36 ± 10% fio.latency_1000ms%
0.16 ± 11% -0.1 0.03 ± 25% -0.1 0.02 ± 8% fio.latency_100ms%
5.59 ± 3% -0.0 5.59 ± 9% +5.7 11.26 ± 15% fio.latency_10ms%
38.25 +1.7 39.92 ± 3% +4.7 42.97 fio.latency_20ms%
0.29 ± 4% -0.1 0.16 ± 2% -0.2 0.06 ± 14% fio.latency_250ms%
0.58 ± 17% +0.1 0.66 ± 10% -0.5 0.08 ± 6% fio.latency_250us%
0.17 ± 43% +0.1 0.24 ± 34% +0.2 0.40 ± 30% fio.latency_2ms%
44.24 -4.3 39.95 -4.2 39.99 ± 3% fio.latency_4ms%
0.03 ± 4% +0.2 0.19 ± 3% +0.1 0.16 ± 10% fio.latency_500ms%
8.56 ± 4% +2.7 11.30 ± 8% -6.2 2.35 ± 5% fio.latency_50ms%
0.03 ± 24% +0.2 0.28 +0.4 0.45 ± 2% fio.latency_750ms%
0.11 ± 38% -0.0 0.10 ± 31% +0.3 0.41 ± 33% fio.latency_750us%
13682713 -9.0% 12457118 -6.7% 12764827 fio.workload
266.36 -8.9% 242.70 -6.5% 248.99 fio.write_bw_MBps
19966634 +3.1% 20578304 -9.2% 18131626 fio.write_clat_90%_us
21976405 +3.2% 22675456 -12.3% 19267584 fio.write_clat_95%_us
92798976 ± 5% +124.1% 2.08e+08 ± 3% +375.3% 4.411e+08 ± 2% fio.write_clat_99%_us
16353304 +9.7% 17942942 +7.0% 17500789 fio.write_clat_mean_us
76828732 +0.4% 77132911 +6.9% 82135424 ± 2% fio.write_clat_stddev
68188 -8.9% 62131 -6.5% 63740 fio.write_iops
523406 +9.7% 574400 +7.0% 560260 fio.write_slat_mean_us
13649919 -3.8% 13133471 +7.9% 14729703 fio.write_slat_stddev

\
 
 \ /
  Last update: 2020-04-13 10:22    [W:0.051 / U:0.108 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site