network performance

[Stefan Lambrev]

Greetings,

Stefan Lambrev wrote:
> Stefan Lambrev wrote:
>> Andrew Thompson wrote:
>>> On Mon, Feb 04, 2008 at 05:26:35PM +0200, Stefan Lambrev wrote:
>>>  
>>>> Greetings,
>>>>
>>>> In my desire to increase network throughput, and to be able to 
>>>> handle more then ~250-270kpps
>>>> I started experimenting with lagg and link aggregation control 
>>>> protocol (lacp).
>>>> To my surprise this doesn't increase the amount of packets my 
>>>> server can handle
>>>>
>>>> Using lagg doesn't improve situation at all, and also errors are 
>>>> not reported.
>>>> Also using lagg increased content switches:
>>>>
>>>> Top showed for CPU states +55%   system, which is quite high?
>>>>
>>>> I'll use hwpmc and lock_profiling to see where the kernel spends 
>>>> it's time.
>>>>     
>>>
>>> Thanks for investigating this. One thing to note is that ip flows from
>>> the same connection always go down the same interface, this is because
>>> Ethernet is not allowed to reorder frames. The hash uses
>>> src-mac, dst-mac, src-ip and dst-ip (see lagg_hashmbuf), make sure when
>>> performance testing that your traffic varies in these values. Adding
>>> tcp/udp ports to the hashing may help.
>>>   
>> The traffic, that I generate is with random/spoofed src part, so it 
>> is split between interfaces for sure :)
>>
>> Here you can find results when under load from hwpmc and lock_profiling:
>> http://89.186.204.158/lock_profiling-lagg.txt
>> http://89.186.204.158/lagg-gprof.txt
>>
> http://89.186.204.158/lagg2-gprof.txt I forget this file :)
>
I found that MD5Transform aways uses ~14% (with rx/txcsum enabled or 
disabled).
And when using without lagg MD5Transform pick up to 20% of the time.
Is this normal?

-- 

Best Wishes,
Stefan Lambrev
ICQ# 24134177