[c-nsp] MTU / BGP

[Gert Doering]

Hi,

On Tue, Jul 21, 2015 at 12:25:59PM +0200, Mark Tinka wrote:
> On 21/Jul/15 09:16, Gert Doering wrote:
> > In that case, your L3 MTU on your routers needs to be set to something
> > lower or equal to the lowest L2 MTU that your packets might hit.
> 
> I generally don't set protocol MTU, only Layer 2 MTU. Normally, your
> protocols should be able to infer their MTU from the Layer 2 MTU.
> 
> But yes, in some cases, setting the Layer 3 MTU may be necessary.

Whatever works for your platforms :-) - if you do L3 subinterfaces on
classic IOS, for example, you can't do physical MTU on subifs, and when
different VLAN/.1q subifs end up with different MTU values, you need
to go for protocol MTU... 

But right, if you can just set physical MTU to cover all available L3
protocols, this is much easier to get right for everything...


[..]
> > Best, get your provider to guarantee you a given value, and regularily
> > *verify* that.  What we do: ping all routers behind "dubious" links with
> > 10k packets - on the way *to* the router, the packet is fragmented to
> > 1500 bytes (because the management system doesn't have a bigger uplink),
[..]
> 
> That's a good workaround.

I try to avoid being bitten by the same "assumption" twice... :-)

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert at greenie.muc.de
fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 291 bytes
Desc: not available
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20150721/0ca65469/attachment.sig>