Re: BUG REPORT: Massive performance drop in routing throughput with 2.4.21

From: Martin Zwickel (martin.zwickel@technotrend.de)
Date: Mon Jun 16 2003 - 08:10:35 EST

On my Home-Gateway I have the same problem.
(route 1-1 with iptables)
But I don't know if it's a broken network cable or something else.

Hmm, but before I switched the kernel from 2.0.x to 2.4.21 I thought I had a
better throughput. (my HUB always gives a few collisions, but it was ok)

i have 2 RTL8139 cards:
00:09.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL-8139/8139C/8139C+ (rev 10)
00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL-8139/8139C/8139C+ (rev 10)

and i still don't know what the carrier means:
eth0 Link encap:Ethernet HWaddr 00:50:BF:1A:02:01
          inet addr:192.168.0.1 Bcast:192.168.0.255 Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
          RX packets:19971100 errors:289699 dropped:1 overruns:20 frame:0
          TX packets:18045589 errors:108785 dropped:0 overruns:0 carrier:217570
          collisions:46041 txqueuelen:100
          RX bytes:2036864361 (1942.5 Mb) TX bytes:808600381 (771.1 Mb)
          Interrupt:11 Base address:0xdf00

eth1 Link encap:Ethernet HWaddr 00:30:84:3D:A3:00
          inet addr:10.0.0.1 Bcast:10.0.0.255 Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
          RX packets:18179395 errors:110 dropped:0 overruns:0 frame:0
          TX packets:19959609 errors:0 dropped:0 overruns:0 carrier:0
          collisions:103850 txqueuelen:100
          RX bytes:1069462782 (1019.9 Mb) TX bytes:2154547542 (2054.7 Mb)
          Interrupt:5 Base address:0xfe00

On Mon, 16 Jun 2003 14:51:35 +0200
Stephan von Krawczynski <skraw@ithnet.com> bubbled:

> On Mon, 16 Jun 2003 14:47:15 +0200 (CEST)
> Maciej Soltysiak <solt@dns.toxicfilms.tv> wrote:
>
> > > there seems to be a real serious problem with 2.4.21, routing through two
> > > ethernet-devices. After 24 hours of a routing-only box the throughput from
> > > ethernet a to ethernet b decreased to something around 4-100 kByte/sec
> > > (100 Mbit network 2 cards ns83820). I had to drop using 2.4.21 on this box
> > > because of this. 2.4.20 is flawless on the machine and ran for around 100
> > > days before without any troubles. Going back to 2.4.20 cured it.
> > Are you using any netfilter patch-o-matic patches?
>
> No.
>
> > Does it also affect eg. ssh latency even on LAN ?
>
> Not very seriously.
>
> > Since I switched my processor from Intel do Amd, I have been experiencing
> > similar but after longer periods of time than yours.
> >
> > I am also using VIA chipset, maybe it's a hardware driver problem.
>
> No, my two boxes have differing mb's. One is VIA, the other is Intel440BX.
>
> But I can also verify that the problem arises with 2.4.21-rc7, too. I guess it
> is effectively an older problem that has not been recognised so far.
>
> > Regards,
> > Maciej
>
> Regards,
> Stephan
>
> -
> 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/

Regards,
Martin 

-- 
MyExcuse:
The static electricity routing is acting up...

Martin Zwickel <martin.zwickel@technotrend.de>
Research & Development

TechnoTrend AG <http://www.technotrend.de>
- 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/

This archive was generated by hypermail 2b29 : Mon Jun 23 2003 - 22:00:18 EST