[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#549018: marked as done (linux-image-2.6-686: Network unresponsive from certain hosts)

Your message dated Tue, 9 Feb 2010 23:10:51 +0100
with message-id <20100209221051.GA6766@galadriel.inutil.org>
and subject line Re: Bug#549018: linux-image-2.6-686: Network unresponsive from certain hosts
has caused the Debian Bug report #549018,
regarding linux-image-2.6-686: Network unresponsive from certain hosts
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)


-- 
549018: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=549018
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message --- 
Package: linux-image-2.6-686
Severity: important

I apologise that I have little information to go on with this bug, I am continuing to attempt to debug it but in a mixed network environment it is quite tricky.

This bug occured while I was testing link aggregation/link bonding in debian. I have 3 ethernet interfaces bound to bond0, their manufacturer and driver information is below:
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)
00:12.0 Ethernet controller: VIA Technologies, Inc. VT6102 [Rhine-II] (rev 78)

I am also using iperf on two computers connected to the same switch in order to attempt to push > 100mbit of traffic across the machine, to check link aggregation / the hash method.

All machines are on 192.168.101.0/24, but the two test machines have static routes to reach each other through the machine with the bonded links. It also has icmp redirects disabled to force it to pass the packets through.

Whenever I initiate a TCP iperf connection, moments later the server side of the connection loses all contact with the link aggregated machine, it will not respond to ping or any other type of traffic

There are no other relevant configuration parameters, the link aggregation box can actually send traffic out, I can see pings hitting the test server and being returned, but they never appear on the link aggregated machine. I can also see no issue with the routing table (network + gateway, nothing special) nor the arp table, nor anything else at all.

Please let me know if I have made some trivial error, but I believe this to be a genuine bug.

-- System Information:
Debian Release: 5.0.3
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.26-2-686 (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

--- End Message ---
--- Begin Message --- 
On Wed, Sep 30, 2009 at 11:20:46PM +0100, Ben Hutchings wrote:
> On Wed, 2009-09-30 at 14:05 +0100, Paul Robins wrote:
> > Package: linux-image-2.6-686
> > Severity: important
> > 
> > I apologise that I have little information to go on with this bug, I
> > am continuing to attempt to debug it but in a mixed network
> > environment it is quite tricky.
> 
> You should report kernel bugs against linux-image-$(uname -r) and not
> against linux-image-2.6-686.  That will automatically include some
> useful system information.
> 
> > This bug occured while I was testing link aggregation/link bonding in
> > debian. I have 3 ethernet interfaces bound to bond0, their
> > manufacturer and driver information is below:
> > 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)
> > 00:12.0 Ethernet controller: VIA Technologies, Inc. VT6102 [Rhine-II] (rev 78)
> > 
> > I am also using iperf on two computers connected to the same switch in
> > order to attempt to push > 100mbit of traffic across the machine, to
> > check link aggregation / the hash method.
> > 
> > All machines are on 192.168.101.0/24, but the two test machines have
> > static routes to reach each other through the machine with the bonded
> > links. It also has icmp redirects disabled to force it to pass the
> > packets through.
> > 
> > Whenever I initiate a TCP iperf connection, moments later the server
> > side of the connection loses all contact with the link aggregated
> > machine, it will not respond to ping or any other type of traffic
> > 
> > There are no other relevant configuration parameters, the link
> > aggregation box can actually send traffic out, I can see pings hitting
> > the test server and being returned, but they never appear on the link
> > aggregated machine. I can also see no issue with the routing table
> > (network + gateway, nothing special) nor the arp table, nor anything
> > else at all.
> 
> Check the MAC stats in /proc/net/dev.  Are the ping responses added to
> any of the Receive counters there?
> 
> > Please let me know if I have made some trivial error, but I believe
> > this to be a genuine bug.
> 
> Please send the output of the following commands after this problem
> occurs:
>     dmesg
>     ip addr show
>     ip link show
>     grep -r . /sys/class/net/bond*

No further feedback, closing the bug. If this can still be reproduced
please reopen and provide the requested information.

Cheers,
         Moritz

--- End Message ---
Reply to: