Bug#699269: IPv6 slow and hangs over failed bonding link

[Gasha <gasha@pie-dabas.net>]

On 01/30/2013 01:09 AM, Jonathan Nieder wrote:
> Hi Gasha,
>
> Gasha wrote:
>
>    
> > Package: linux
> > Version: 3.2.0-2-amd64
> >      
> What version?  The above is part of the package name, not the version
> number --- you can see the currently running version of Linux with
>
> 	cat /proc/version
>
> (look in the parentheses) or the currently installed version with
>
> 	dpkg-query -W linux-image-$(uname -r)
>
> ([1] has more details.)
>    

Host:

*$ cat /proc/version
Linux version 3.2.0-2-amd64 (Debian 3.2.12-1) 
(debian-kernel@lists.debian.org) (gcc version 4.6.3 (Debian 4.6.3-1) ) 
#1 SMP Tue Mar 20 18:36:37 UTC 2012
*
Guest VM:

# cat /proc/version
Linux version 3.2.0-3-amd64 (Debian 3.2.23-1) 
(debian-kernel@lists.debian.org) (gcc version 4.6.3 (Debian 4.6.3-8) ) 
#1 SMP Mon Jul 23 02:45:17 UTC 2012


> [...]
>    
> > Two ethernet interfaces configured in bond0: (/etc/network/interfaces)
> >      
> [...]
>    
> > One ethernet cable was disconnected, before planned reconfiguration
> > works: (dmesg)
> >
> > [21698247.247121] bonding: bond0: link status down for interface eth1, disabling it in 200 ms.
> > [21698247.446804] bonding: bond0: link status definitely down for interface eth1, disabling it
> >
> > ifconfig eth1 still shows it up.
> > HP Procurve Gigabit switch configured as multilink trunk.
> > Everything worked when both cables were connected.
> >
> > VirtualBox machine is running using bridged interface br576
> >
> > I tried to copy large file over NFS4 / IPv6 and it freeze to very
> > low speed. cp command hung, and we had to reboot virtual machine, as
> > it was hanging. at destination NFS server file was growing but
> > slowly.
> > tcpdump showed short bursts of data and then again 5-15 seconds nothing.
> >
> > After reboot i reproduced this behaviour using "scp":
> > scp to ipv6 address:
> >      
> I'll let others more knowledgeable than I am answer about the
> substance of your report. :)
>
> Is this a regression (if so, do you remember which upgrade broke
> things?) or has this procedure always failed this way?  Are more
> recent 3.2.y kernels and 3.7.y from experimental also affected?
>    

Well, this is production server. But i have similar setup machines, so 
that it is possible to repeat test and try other kernels also. I'm 
trying to migrate infrastructure so that core services work IPv6 only.

> Thanks for writing,
> Jonathan
>
> [1] http://kernel-handbook.alioth.debian.org/ch-versions.html#s-version-types
>