[ovs-discuss] Questions about some performance issues / bandwitch limitation inside openvswitch

[Jesse Gross]

On Mon, Jan 24, 2011 at 4:55 AM, Benoit ML <ben42ml at gmail.com> wrote:
> Hi,
> Thank you for your reply.
> In bridge mode performance is good but I don't have done deep testing ...
> Well you pointed me the mtu and effectively, it seems that I have got issue
> about that.
> It seems that the mtu of the gre tunnel is 1524. Can you confirm that ?
> Well have you some suggestion for the guest MTU ? and for the KVM tap device
> ?

Where does 1524 come from?  What is the underlying MTU of your
network?  GRE has some overhead for the encapsulation header, which
will reduce your effective MTU.  The header is variable sized based on
what options are enabled but by default on a network with a 1500 byte
MTU the tunnel MTU should be 1462.  The Open vSwitch implementation
attempts to handle path MTU discovery both inside and outside the
tunnel to provide the optimal MTU to the sender without any additional
configuration required.  However, the exact behavior depends on the OS
of the sender and its configuration.

If an incorrect MTU is chosen it can result in severe performance problems.