[prev in list] [next in list] [prev in thread] [next in thread] 

List:       e1000-devel
Subject:    Re: [E1000-devel] RSS on 82576
From:       Luca Deri <deri () ntop ! org>
Date:       2009-11-20 17:44:12
Message-ID: DB845EE9-CC38-47BF-9567-BBFD91209015 () ntop ! org
[Download RAW message or body]

Jesse
suppose you have a TCP connection A:b <-> B:b (A = IP, b = port). On 82576 A:a -> B:b \
goes to queue X, and B:b -> A:a goes to queue Y, where X != Y, whereas I would like \
to have X == Y.

This looks normal to me according to the specs \
(http://download.intel.com/design/network/datashts/82576_Datasheet.pdf) as the \
hashing function (page 246) is not producing the same result for both directions.

I was wondering if I can force RSS to put packets of the same TCP connection onto the \
same queue as otherwise received packets are shuffled (this is the problem I'm \
willing to solve).

I use igb 2.0.6 and I insert the igb module as "insmod ./igb.ko VMDQ=2,2 RSS=2,2" on \
my Core2Duo machine.

Thanks for your help.

Luca

On Nov 20, 2009, at 6:35 PM, Brandeburg, Jesse wrote:

> On Fri, 20 Nov 2009, Luca Deri wrote:
> > I would like to force RSS to balance packets from the same TCP 
> > connection to be sent on the same queue. Today RSS splits them into two 
> > different queues.
> > 
> > How can I do that?
> 
> Luca, I'm a little confused.  If it is the same TCP connection by 
> definition it will come to the same RSS queue.  Thats what RSS is for.
> 
> Are you talking about after a reboot?


------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel


[prev in list] [next in list] [prev in thread] [next in thread]