[Bridge] Bridging over a bonded interface?

[Bridge] Bridging over a bonded interface?

[Ravi Ramamirtham]

[-- Attachment #1: Type: text/plain, Size: 470 bytes --]

Hi,
  This may sound crazy but I will go ahead anyway. Here's my scenario:

ifconfig bond0 10.10.10.1 netmask 255.0.0.0
ifenslave bond0 eth0
ifenslave bond0 eth1

brctl addbr stp0
brctl addif stp0 bond0
brctl stp stp0 on

Is this a valid scenario? I am trying to bond two interfaces and run STP
over the bonded interface..The problem I am trying to solve is to run STP
over aggregated links between two nodes. Any suggestions/pointers would be
much appreciated.

Thanks

[-- Attachment #2: Type: text/html, Size: 587 bytes --]

Re: [Bridge] Bridging over a bonded interface?

[Stephen Hemminger]

On Mon, 16 Apr 2007 17:10:37 -0400
"Ravi Ramamirtham" <rrshanks@gmail.com> wrote:

> Hi,
>   This may sound crazy but I will go ahead anyway. Here's my scenario:
> 
> ifconfig bond0 10.10.10.1 netmask 255.0.0.0
> ifenslave bond0 eth0
> ifenslave bond0 eth1
> 
> brctl addbr stp0
> brctl addif stp0 bond0
> brctl stp stp0 on
> 
> Is this a valid scenario? I am trying to bond two interfaces and run STP
> over the bonded interface..The problem I am trying to solve is to run STP
> over aggregated links between two nodes. Any suggestions/pointers would be
> much appreciated.
> 
> Thanks

The Linux code won't mind, but if you are bonding into another vendor's
switch, it is probably going to get horribly confused.

Re: [Bridge] Bridging over a bonded interface?

[Axel Thimm]

[-- Attachment #1: Type: text/plain, Size: 1339 bytes --]

On Mon, Apr 16, 2007 at 05:10:37PM -0400, Ravi Ramamirtham wrote:
> Hi,
>  This may sound crazy but I will go ahead anyway. Here's my scenario:
> 
> ifconfig bond0 10.10.10.1 netmask 255.0.0.0
> ifenslave bond0 eth0
> ifenslave bond0 eth1
> 
> brctl addbr stp0
> brctl addif stp0 bond0
> brctl stp stp0 on
> 
> Is this a valid scenario? I am trying to bond two interfaces and run STP
> over the bonded interface..The problem I am trying to solve is to run STP
> over aggregated links between two nodes. Any suggestions/pointers would be
> much appreciated.

Just happened that I asked the same 9 days ago:

http://lists.linux-foundation.org/pipermail/bridge/2007-April/001764.html

Depending on the switch on the other side of the wire, you will get
the arp packets you send out on the other link of the bond. This will
make an outgoing arp packet register on the internal port of the
bridge and reregister a couple of ms later on the outer port.

This leads to packets loss of > 99%, e.g. a 2x1GB bond degarded to
70kB/sec.

It would be nice if the bridging code could prevent this. Either by
identifying the outgoing arp packet that immediately returns and not
updating the forwarding database, or allowing to hardcode some static
arp entries into the forwarding database.
-- 
Axel.Thimm at ATrpms.net

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [Bridge] Bridging over a bonded interface?

[Ravi Ramamirtham]

Thanks Axel and Stephen. I am planning to use Linux at
both ends. I didn't get a chance to verify the setup
today but will try and get something going tomorrow.


--- Axel Thimm <Axel.Thimm@atrpms.net> wrote:

> On Mon, Apr 16, 2007 at 05:10:37PM -0400, Ravi
> Ramamirtham wrote:
> > Hi,
> >  This may sound crazy but I will go ahead anyway.
> Here's my scenario:
> > 
> > ifconfig bond0 10.10.10.1 netmask 255.0.0.0
> > ifenslave bond0 eth0
> > ifenslave bond0 eth1
> > 
> > brctl addbr stp0
> > brctl addif stp0 bond0
> > brctl stp stp0 on
> > 
> > Is this a valid scenario? I am trying to bond two
> interfaces and run STP
> > over the bonded interface..The problem I am trying
> to solve is to run STP
> > over aggregated links between two nodes. Any
> suggestions/pointers would be
> > much appreciated.
> 
> Just happened that I asked the same 9 days ago:
> 
>
http://lists.linux-foundation.org/pipermail/bridge/2007-April/001764.html
> 
> Depending on the switch on the other side of the
> wire, you will get
> the arp packets you send out on the other link of
> the bond. This will
> make an outgoing arp packet register on the internal
> port of the
> bridge and reregister a couple of ms later on the
> outer port.
> 
> This leads to packets loss of > 99%, e.g. a 2x1GB
> bond degarded to
> 70kB/sec.
> 
> It would be nice if the bridging code could prevent
> this. Either by
> identifying the outgoing arp packet that immediately
> returns and not
> updating the forwarding database, or allowing to
> hardcode some static
> arp entries into the forwarding database.
> -- 
> Axel.Thimm at ATrpms.net
> > _______________________________________________
> Bridge mailing list
> Bridge@lists.linux-foundation.org
>
https://lists.linux-foundation.org/mailman/listinfo/bridge
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com