Linux Routing Performance inferior?

Linux Routing Performance inferior?

[Ram Chandar]

Quoted from a recent mail to freebsd mailing list.

"FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
Linux can't do much more than 100kpps"

http://lists.freebsd.org/pipermail/freebsd-net/2004-September/004840.html

Is this indeed the case?

Ram Chandar.
--

Re: Linux Routing Performance inferior?

[William Stearns]

Good afternoon, Ram,

On Wed, 8 Sep 2004, Ram Chandar wrote:

> Quoted from a recent mail to freebsd mailing list.
> 
> "FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
> Linux can't do much more than 100kpps"
> 
> http://lists.freebsd.org/pipermail/freebsd-net/2004-September/004840.html
> 
> Is this indeed the case?

	I'm sure others here have far better examples, but one post to the 
netfilter-devel list last December provided an example of a firewall that 
could process 580kpps with netfilter/conntrack turned off.  Granted, the 
post noted that adding netfilter brought that down to 450kpps, and adding 
conntrack on top of that brought it down to 295kpps, but all three of 
those numbers are well over the claimed 100kpps.
	Cheers,
	- Bill

---------------------------------------------------------------------------
        "While it may be true that a watched pot never boils, the one
you don't keep an eye on can make an awful mess of your stove."
        -- Edward Stevenson 
(Courtesy of Slashdot)
--------------------------------------------------------------------------
William Stearns (wstearns@pobox.com).  Mason, Buildkernel, freedups, p0f,
rsync-backup, ssh-keyinstall, dns-check, more at:   http://www.stearns.org
--------------------------------------------------------------------------

Re: Linux Routing Performance inferior?

[Nathan Bryant]

William Stearns wrote:
> Good afternoon, Ram,
> 
> On Wed, 8 Sep 2004, Ram Chandar wrote:
> 
> 
>>Quoted from a recent mail to freebsd mailing list.
>>
>>"FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
>>Linux can't do much more than 100kpps"
>>
>>http://lists.freebsd.org/pipermail/freebsd-net/2004-September/004840.html
>>
>>Is this indeed the case?
> 
> 
> 	I'm sure others here have far better examples, but one post to the 
> netfilter-devel list last December provided an example of a firewall that 
> could process 580kpps with netfilter/conntrack turned off.  Granted, the 
> post noted that adding netfilter brought that down to 450kpps, and adding 
> conntrack on top of that brought it down to 295kpps, but all three of 
> those numbers are well over the claimed 100kpps.

Nonetheless, FreeBSD has some advantages. They achieved their results 
using a fast forwarding path (enabled via sysctl) that processes 
forwarded packets to completion entirely within the interrupt handler.

Re: Linux Routing Performance inferior?

[Tomasz Torcz]

On Wed, Sep 08, 2004 at 02:41:34PM -0400, Nathan Bryant wrote:
> >>"FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
> >>Linux can't do much more than 100kpps"
> 
> Nonetheless, FreeBSD has some advantages. They achieved their results 
> using a fast forwarding path (enabled via sysctl) that processes 
> forwarded packets to completion entirely within the interrupt handler.

 I've already posted presentation about those features (*) to netdev.
Some ideas looks interesting enough to be implemented in Linux.

* http://people.freebsd.org/~andre/FreeBSD-5.3-Networking.pdf

-- 
Tomasz Torcz                        To co nierealne - tutaj jest normalne.
zdzichu@irc.-nie.spam-.pl          Ziomale na życie mają tu patenty specjalne.

Re: Linux Routing Performance inferior?

[Norbert van Nobelen]

BSD is known for good network performance, however I don't know benchmarks. 
I think the difference is to big: The routing/IP stack combined being 10 times 
less efficient is too much.

They also don't mention which linux kernel they use. Reading the 
FreeBSD-5.3-Networking.pdf they did some optimasations which are probably not 
advisable if you don't use your box as a router.
The goal of this person is as far as I can see to build a router only, so in 
theory you could build in the same optimasations in network stack of linux

Also look at page 11: The fastforwarding is a solid positive step on how a 
router should work. So even the performance of FreeBSD is not considered like 
a real router OS.



On Wednesday 08 September 2004 19:36, you wrote:
> Quoted from a recent mail to freebsd mailing list.
>
> "FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
> Linux can't do much more than 100kpps"
>
> http://lists.freebsd.org/pipermail/freebsd-net/2004-September/004840.html
>
> Is this indeed the case?
>
> Ram Chandar.
> --
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: Linux Routing Performance inferior?

[Matt Kavanagh]

On Wed, Sep 08, 2004 at 11:06:17PM +0530, Ram Chandar wrote:
> 
> Quoted from a recent mail to freebsd mailing list.
> 
> "FreeBSD (5.x) can route 1Mpps on a 2.8G Xeon while
> Linux can't do much more than 100kpps"
> 
> http://lists.freebsd.org/pipermail/freebsd-net/2004-September/004840.html
> 
> Is this indeed the case?

Seems to be pretty much just biased conjecture IMO. I wouldn't
dismiss the possibility of FreeBSD having (in some situations)
significantly better routing performance than linux in the same
situation..but getting me to believe that would require proper,
objective benchmarks.

All from a user's perspective.