[Lxc-users] Poor network performance between container and host

[Dominik Schulz]

Am Sonntag 31 Januar 2010 22:59:05 schrieben Sie:
> Dominik Schulz wrote:
> > Hi,
> > I'm currently evaluating LXC as an alternative to other contextualisation
> > technologies, and I have the problem that network connections between the
> > container and the host on which it is running are very slow (about 35
> > kbit/s) wheras connections to other system can exploit the full
> > connection speed.
> > Is this a known issue or is someting wrong with my configuration?
> > The relevant snippet of my lxc config:
> > lxc.network.type = veth
> > lxc.network.flags = up
> > lxc.network.link = br0
> > lxc.network.name = eth0
> > lxc.network.mtu = 1500
> > My Bridge (br0) has an MTU of 1480, could that be the problem?
> Having a different mtu size will contribute to decrease the
> performances. But the difference is too big to be related only to the
> mtu IMO.
> Are you sure there is not another container somewhere with the same
> network configuration messing the communication ? Perhaps on another host ?
I'm pretty sure that the network address is unique in my subnet. But I see a 
lot of TCP Retransmisions (about 50% of the packages!) and duplicated TCP ACKs 
in the packet dump on the containers interface. This is very strange.

-- 
Mit freundlichen Grüßen / Best Regards
Dominik
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.linuxcontainers.org/pipermail/lxc-users/attachments/20100202/4ca297b9/attachment.pgp>