Possible evidence of performance regression for 8.1-S (vs. 7.1)
David Wolfskill
david at catwhisker.org
Fri Oct 29 16:58:53 UTC 2010
On Tue, Oct 26, 2010 at 10:34:08AM -0400, Mike Tancsa wrote:
> ...
> ULE is the default in 7 as well. Perhaps remove some of the kernel
> options not in 7, that are in 8 by default? What is the disk
> subsystem ? just ata ?
>
> They seem innocuous enough, but worth a try
>
> options HWPMC_HOOKS # Necessary kernel hooks for hwpmc(4)
> options MAC # TrustedBSD MAC Framework
> options FLOWTABLE # per-cpu routing cache
OK; I now have the results from testing that.
As noted earlier, we were using the MAC kernel config in 7.1, so I did
not test removing the MAC option.
I did build a kernel removing each of HWPMC_HOOKS and FLOWTABLE; results
will be shown below as being associated with os "8.1-S-". To reduce the
need for folks to find earlier messages to see the previous results,
I'll paste them in, too:
start stop real user sys host os
1287525363 1287546846 21483.13 82628.20 21703.09 ref-7x 7.1-R+
1287548005 1287569100 21094.63 82853.19 22185.02 ref-7x 7.1-R+
1287570300 1287591371 21071.33 82756.81 21943.22 ref-7x 7.1-R+
1287592592 1287614103 21511.23 82637.30 21849.90 ref-7x 7.1-R+
1287615323 1287636770 21446.42 82715.81 21708.97 ref-7x 7.1-R+
1287436357 1287461948 25590.99 81502.22 18115.07 ref-8x 8.1-S at r214029
1287462797 1287488766 25969.26 81452.14 17920.14 ref-8x 8.1-S at r214029
1287489641 1287515287 25645.84 81548.40 18256.52 ref-8x 8.1-S at r214029
1287516151 1287541481 25329.64 81546.23 18294.10 ref-8x 8.1-S at r214029
1287542355 1287568599 26244.59 81431.47 17902.39 ref-8x 8.1-S at r214029
1287710312 1287732046 21733.20 82688.01 22108.95 ref-8x 7.1-R+
1287733360 1287754549 21188.88 82869.09 21890.83 ref-8x 7.1-R+
1287755881 1287777566 21684.09 82772.50 21933.74 ref-8x 7.1-R+
1287879508 1287905173 25665.03 81696.22 18278.42 ref-8x 8.1-S[7.1-R+ user]
1287906039 1287931709 25669.48 81735.00 18265.78 ref-8x 8.1-S[7.1-R+ user]
1287932573 1287958275 25700.99 81700.40 18246.04 ref-8x 8.1-S[7.1-R+ user]
1287959144 1287984859 25714.74 81800.40 18346.57 ref-8x 8.1-S[7.1-R+ user]
1287985748 1288011752 26004.33 81569.28 17967.68 ref-8x 8.1-S[7.1-R+ user]
1288198095 1288223776 25681.17 81550.12 18442.21 ref-8x 8.1-S-
1288224673 1288250350 25676.86 81561.61 18448.39 ref-8x 8.1-S-
1288269106 1288294837 25730.45 81535.38 18396.07 ref-8x 8.1-S-
1288295688 1288321338 25649.35 81552.34 18424.28 ref-8x 8.1-S-
[Only 4 results posted, as one of the attempted builds failed.]
And here are some ministat plots. As we have 3 "environment"
variables (which machine is used, which kernel, and which userland),
as well as 3 distinct variables being measured (real [wall clock]
time; user CPU seconds; system CPU seconds), the file names reflect
the measured variable ("real", "user", or "sys"), a "_h" to indicate
that a "host designation" ("7" or "8") follows, a "_k" to indicate
that a kernel designation follows ("7", "8", or "8-"), and a "_u" to
indicate that a userland desgnation follows ("7" or "8").
So -- real time first:
dwolf-bsd(8.1-S)[30] ministat -s real*
x real_h7_k7_u7
+ real_h8_k78_u7
* real_h8_k7_u7
: real_h8_k8-_u8
# real_h8_k8_u8
+------------------------------------------------------------------------------+
|x * xx ** # #O@ O #|
||___A_M| |
| |MA_| |
| |____A_M_| |
| |A |
| |___M_A____| |
+------------------------------------------------------------------------------+
N Min Max Median Avg Stddev
x 5 21071.33 21511.23 21446.42 21321.348 218.96441
+ 5 25665.03 26004.33 25700.99 25750.914 143.20282
Difference at 95.0% confidence
4429.57 +/- 269.817
20.7753% +/- 1.26548%
(Student's t, pooled s = 185.003)
* 3 21188.88 21733.2 21684.09 21535.39 301.08941
No difference proven at 95.0% confidence
: 4 25649.35 25730.45 25679.015 25684.457 33.745994
Difference at 95.0% confidence
4363.11 +/- 264.927
20.4636% +/- 1.24254%
(Student's t, pooled s = 166.989)
# 5 25329.64 26244.59 25645.84 25756.064 355.43859
Difference at 95.0% confidence
4434.72 +/- 430.527
20.7994% +/- 2.01923%
(Student's t, pooled s = 295.197)
dwolf-bsd(8.1-S)[31]
User CPU:
dwolf-bsd(8.1-S)[31] ministat -s user*
x user_h7_k7_u7
+ user_h8_k7_u7
* user_h8_k8-_u8
: user_h8_k8_u7
# user_h8_k8_u9
+------------------------------------------------------------------------------+
| O :|
|x* O :|
|** #O# :|
|A| |
||A |
| A |
| A|
| |A |
+------------------------------------------------------------------------------+
N Min Max Median Avg Stddev
x 5 21071.33 21511.23 21446.42 21321.348 218.96441
+ 3 21188.88 21733.2 21684.09 21535.39 301.08941
No difference proven at 95.0% confidence
* 4 25649.35 25730.45 25679.015 25684.457 33.745994
Difference at 95.0% confidence
4363.11 +/- 264.927
20.4636% +/- 1.24254%
(Student's t, pooled s = 166.989)
: 5 81569.28 81800.4 81700.4 81700.26 84.271865
Difference at 95.0% confidence
60378.9 +/- 241.959
283.185% +/- 1.13482%
(Student's t, pooled s = 165.902)
# 5 25329.64 26244.59 25645.84 25756.064 355.43859
Difference at 95.0% confidence
4434.72 +/- 430.527
20.7994% +/- 2.01923%
(Student's t, pooled s = 295.197)
dwolf-bsd(8.1-S)[32]
Finally, system CPU:
dwolf-bsd(8.1-S)[32] ministat -s sys*
x sys_h7_k7_u7
+ sys_h8_k7_u7
* sys_h8_k8-_u8
: sys_h8_k8_u7
# sys_h8_k8_u8
+------------------------------------------------------------------------------+
|# : ** x + |
|#: # ##:** x x+x +x|
| |__A___| |
| |MA_| |
| AM |
| |__AM| |
||___A__| |
+------------------------------------------------------------------------------+
N Min Max Median Avg Stddev
x 5 21703.09 22185.02 21849.9 21878.04 199.04094
+ 3 21890.83 22108.95 21933.74 21977.84 115.55385
No difference proven at 95.0% confidence
* 4 18396.07 18448.39 18433.245 18427.737 23.457507
Difference at 95.0% confidence
-3450.3 +/- 239.945
-15.7706% +/- 1.09674%
(Student's t, pooled s = 151.242)
: 5 17967.68 18346.57 18265.78 18220.898 146.5163
Difference at 95.0% confidence
-3657.14 +/- 254.882
-16.716% +/- 1.16501%
(Student's t, pooled s = 174.763)
# 5 17902.39 18294.1 18115.07 18097.644 182.87428
Difference at 95.0% confidence
-3780.4 +/- 278.75
-17.2794% +/- 1.27411%
(Student's t, pooled s = 191.129)
dwolf-bsd(8.1-S)[33]
Peace,
david
--
David H. Wolfskill david at catwhisker.org
Depriving a girl or boy of an opportunity for education is evil.
See http://www.catwhisker.org/~david/publickey.gpg for my public key.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-performance/attachments/20101029/5815fe02/attachment.pgp
More information about the freebsd-performance
mailing list