Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
1/ Tests
1.1/ DRBD 0.7.10-4 with 2.6.8-3-686 kernel
Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
taro 3G 20698 75 23434 13 17667 9 29679 90 45123 11 93.1 0
------Sequential Create------ --------Random Create--------
-Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
16 324 25 +++++ +++ 856 7 331 24 +++++ +++ 722 15
1.2/ DRBD 0.7.10-4 with 2.6.15-1-686 kernel
Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
taro 3G 23964 80 24899 13 17821 9 32166 86 45217 9 94.0 0
------Sequential Create------ --------Random Create--------
-Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
16 324 25 +++++ +++ 378 8 327 25 +++++ +++ 296 6
1.3/Analyze
We can see with this test that the kernel really increase the performance for :
-Sequential Output
-Sequential Input
And the kernel also seems to impact delete action (why ???).
I am making direct access bench with 2.6.15-1-686 kernel to confirm our results.
I'll post it in a few hours.
2/ Remaining questions
2.1/ Why do I have difference on reading between tests with and without DRBD whereas it is said on
the site that DRBD has no impact on readings?
2.2/ why the reading per character (Sequential Input Per Chr) with DRBD (both versions) is faster
than without DRBD ?
Is it because the node2's disk is better, the read is made on the 2 nodes and the fastest response
is used?
However I've read on the first page of the DRBD's site :"Reads are always carried out locally."
2.3/ I've read in the drbd-announce mailinglist of May that :
"With drbd-0.7.18 on top of DRBD we got only
about 70MB/sec read performance, while with that modification
in 0.7.19 we can get the whole 320MB/sec out of DRBD as well."
The results between DRBD 0.7.10-4 and DRBD 0.7.19-1 seem a bit similary.
In which conditions, does the read performance increase like this?
Julien WICQUART
Julien WICQUART a écrit :
> 1/ Test of DRBD 0.7.10 with 2.6.15-1-686 kernel.
>
> I will make the test of DRBD 0.7.10 with 2.6.15 kernel.
>
> It will take less than 3 hours on one node, so I will post the result tomorrow.
>
>
> 2/ Test of DRBD 0.7.19 with 2.6.8-3-686 kernel.
>
> I've got a dependences problem with Debian packages. If I want to do so, I need to compile DRBD and
> some other packages.
>
> I think the first test is the most interesting, so I won't make the 2nd test, except if you really
> need it (in this case, please contact me).
>
> Julien WICQUART
>
>
> Lars Ellenberg a écrit :
>> / 2006-07-10 12:09:52 +0200
>> \ Julien WICQUART:
>>> hi,
>>>
>>> i've made some performance benchs with bonnie++ on DRBD 0.7.10-4 and DRBD 0.7.19-1.
>>>
>>> But i don't understand some of my results.
>>>
>>>
>>> 1/ description
>>>
>>> 1.1/ Hardware description
>>> Here are the descriptions of the 2 servers:
>>> common description:
>>> - Intel(R) Celeron(R) CPU 2.40GHz
>>> - 256MB SDRAM
>>> - network card 1Gbps
>>>
>>> node1:
>>> - HD IDE disk 41GB ExcelStor Technology J240
>>>
>>> node2:
>>> - HD IDE disk 41GB IBM IC35L060AVV207-0
>>>
>>> The hard drive of node2 seems to be better than the node1's.
>>>
>>>
>>> 1.2/ Software description
>>>
>>> with DRBD 0.7.10-4:
>>> - debian stable
>>> - kernel 2.6.8-3-686
>>>
>>> with DRBD 0.7.19-1:
>>> - debian testing
>>> - kernel 2.6.15-1-686
>> you did test
>> 0.7.19 with 2.6.8-3-686 and
>> 0.7.10 with 2.6.15 as well?
>>