Re: MSTP backward compatibility ?

Hi Jacek,

Not sure why you concluded that load-balancing is impossible in this case.
From the viewpoint of PVST+ side, the MSTP CIST Root appears to be a root
bridge for every VLAN in the topology. By adjusting STP costs on the links
connecting PVST+ domain to MSTP domain you can select elect root ports on
per-VLAN basis and provide limited form of load-balancing just as usual with
PVST+ scenarios.

-- 
Petr Lapukhov, petr_at_INE.com
CCIE #16379 (R&S/Security/SP/Voice)
CCDE #20100007
Internetwork Expert, Inc.
http://www.INE.com <http://www.ine.com/>
Toll Free: 877-224-8987
Outside US: 775-826-4344
2011/3/1 Jacek <q.192.168.1.0_at_gmail.com>
> Right, root must be on MST, it works now but not fully...
>
> When MST switch sends priority to a PVST switch it sends default priority
> 32768, WITHOUT 12 bits extended system id. This is even the case if MST
> switch is configured as root. It will still send 32768 to a PVST switch.
> This is good because PVST switch will always see an MST switch with lower
> priority than itself or other PVST switches. This is assuming that PVST
> switches do not have priority configured manually.
>
> But...
> in this topology:
>
> SW1=======SW2 ---> MSTP
> ||  \\  // ||
> ||   \\//  ||
> ||   //\\  ||
> || //   \\ ||
> SW3=======SW4 ---> PVST
>
> SW1 and SW2 will both send priority 32768 to SW3 and SW4. As a result SW3
> and SW4 will have to compare MAC addresses to determine the root bridge. In
> my case PVST chooses SW1 is the root for all vlans. Load balancing is then
> impossible.
>
> Is my logic OK ?
>
>
>
> On Tue, Mar 1, 2011 at 3:45 PM, Rares Donca <rares.donca_at_gmail.com> wrote:
>
> > Hello,
> >
> > The spanning-tree root for all Vlans has to be in MSTP, for that to work.
> >
> > I recommend you to read these excellent blog posts who clearly explain
> what
> >
> > is happening there.
> >
> > http://blog.ine.com/2008/07/27/mstp-tutorial-part-i-inside-a-region/
> > http://blog.ine.com/2008/09/24/mstp-tutorial-part-ii-outside-a-region/
> >
> > Regards,
> > Rares
> >
> > On Tue, Mar 1, 2011 at 10:33 PM, Jacek <q.192.168.1.0_at_gmail.com> wrote:
> >
> >> I have a question regarding MSTP backward compatibility with PVST.
> >> According
> >> to Cisco documentation "MSTP maintain backward compatibility with
> >> equipment
> >> that is based on PVST+"
> >>
> >> I am testing this in the lab and and it does not work. In the topology
> >> below
> >> I have MSTP and PVST switches connected. I tried several combination of
> >> stp
> >> priority on different switches and it looks like MST and PVST switches
> >> will
> >> not exchange stp priority at all.
> >>
> >> Here is the scenario:
> >>
> >> SW1=======SW2 ---> MSTP
> >> ||  \\  // ||
> >> ||   \\//  ||
> >> ||   //\\  ||
> >> || //   \\ ||
> >> SW3=======SW4 ---> PVST
> >>
> >> For example, if I configure SW4 to be root bridge for vlan 10 I will end
> >> up
> >> with two switches claiming to be root for vlan 10: SW1 and sw4 will
> claim
> >> to
> >> be root.
> >>
> >> It looks to me that MST and PVST switches do not exchange stp priority,
> >> any
> >> ideas about "backward compatibility" ?
> >>
> >> BTW, I tested a mix of RSTP and PVST+ in the similar topology and it
> >> worked
> >> OK.
> >>
> >>
> >> Blogs and organic groups at http://www.ccie.net
> >>
> >> _______________________________________________________________________
> >> Subscription information may be found at:
> >> http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
Blogs and organic groups at http://www.ccie.net