Hi Rob,
Funny you should mention this, actually. I struggled with the same lab a few months ago and posted a question which was never answered: In V1, L3, 3.8 we are asked to ensure that Cat1 takes specific backup links > to reach the STP root, Cat4. Cat1 has two equal-cost (FE) paths to Cat3, > and two equal-cost FE paths to Cat2, plus the Gig link to Cat2. Cat1 must > prefer the "higher-numbered" port on Cat3, then the "higher-numbered" port > on Cat2 if the links to Cat4 go down. I took that to mean the Fa0/24 link > to Cat2. The restriction is that we can use only one interface command on > Cat1. > > I solved it by tweaking port-priorities on both Cat2 and Cat3 to make Cat1 > prefer the higher ports, and then in order to ensure the port-priority was > used in the port selection for the Cat1-Cat2 link, I used my one interface > command on Cat1 to make the cost of the Gig link equal to the FE links > (200000, this was MST). At that point I had to come up with another way to > make Cat3 the first preferred path, so I sweetened the interface cost on > Cat3's links to Cat4. This worked as desired when both of Cat1's links to > Cat4 were dropped and then *both* links to Cat3 were dropped. > > Now, I missed the fact that the Cat2 link should be used only if that > higher port on Cat3 goes down, not both links on Cat3. So OK, I missed that > one. But in the DSG they use the one interface command on Cat1 to tweak the > port cost on the preferred link from Cat1 to Cat3, preferring it and > leaving the others default. They then did the same priority tweak on Cat2 I > did. OK, I see how that accomplishes the goal, *IF* you ignore the fact > that the Gig link from Cat1-Cat2 would win next due to lower cost with no > consideration for port priority. > > In the DSG they seem to complete ignore the fact that the Gig link > bandwidth between Cat1-Cat2 had to have the cost fixed. Their output for > Cat1's 'sh spann' shows: > > Fa0/1 Desg FWD 200000 > Gi0/2 Desg FWD 200000 > Fa0/4 Desg FWD 200000 > > Which is kinda wacky because the Gig links should be listed last and the > Cat1-Cat2 link is Gig0/1, not Gi0/2. It seems they doctored the output here > and labeled Fa0/2 as Gi0/2 and the DSG makes no mention of Gi0/1. > > Anyone have any guidance on this? It's really very frustrating to see that > the DSG's solutions aren't even feasible within the constraints they > dictate without ignoring or "pre-fixing" part of the lab topology. So, I found the same thing as you... the DSG solution wasn't quite applicable to the topology unless you know what they were trying to do and "fixed" it ahead of time. I frequently found in the switching labs that the Gig link from Cat1 to Cat2 was ignored in the lab setup, but in the DSG there would be a comment like "Let's just knock the gig link down so it doesn't interfere with what we're doing" or it would be clear that the gig link had been disabled in the authors rack. On Fri, Feb 10, 2012 at 2:25 PM, Rob Tyrrell <[email protected]> wrote: > Thanks, Bob! This is extremely helpful. Especially the part about "2. If > root cost through two bridges are equal, use lowest bridge ID." > > Cat4 is actually the root in this topology: > > > Cat1-------------Cat2 > | - - | > | - | > | - - | > Cat3--------------Cat4 > > There are 2g port-channels between each Cat, in a full mesh. The lab has > you shutdown the links between Cat1 and Cat4, so that you have to either go > via Cat2 or Cat3. It has you lower the cost on one of the links from Cat1 > to Cat3, so STP flows through Cat 3, naturally. Then it has you shutdown > the link you lowered the cost on, and the DSG says you SHOULD now be > traversing Cat2, as that is where you enabled a lower port-priority. > > My take: Based on your info, it will never get to this decision, as Cat3 > has a lower BID (MAC, since priorities are both default). So possibly the > DSG is not accurate based on the hardware in my rack. :( > > Thanks also for the info on where to verify port-priority!! :) > > Cheers, > > Rob > > > > > On Feb 10, 2012, at 2:05 PM, Bob McCouch wrote: > > Hi Rob, > > Remember that port priority is only evaluated if you have multiple ports > that connect to the same "upstream" bridge. Root port selection uses the > following preference: > > 1. Lowest root cost > 2. If root cost through two bridges are equal, use lowest bridge ID > 3. If bridge IDs are equal, use lowest port ID (mixture of port number and > port priority) > > In your output, it sounds like Cat3 is the root. In that case, Cat1 won't > care about the priorities advertised by Cat2. If you want Cat2 to be the > root, you need to give it a lower bridge ID (MSTP priority) on the desired > VLANs. If your goal was to make Cat1 go through Cat2 to reach the root > Cat3, you could use port cost, but that would have to be done on Cat1. > > Also note that the "Prio" value shown in "sh spann" does NOT display the > port priority of the incoming BPDUs. It displays the local priority shown > in BPDUs sent OUT from the bridge that displays that output. If you look on > Cat2, you should see the desired priority in "sh spann," or from Cat1 you > can do a "sh spann interface fa0/24 detail" (I think... going from memory) > and if BPDUs are being received on this port you should see the advertised > port priority. > > Hope that helps, > Bob > > > > On Fri, Feb 10, 2012 at 12:04 PM, Rob Tyrrell <[email protected]> wrote: > >> I am having an issue with port-priority influencing the root cost of its >> downstream neighbor. I have configured the port-priority according to the >> DSG on Cat2 Fa0/24: >> >> Cat2-MST(config-if)#do sh run int f0/24 >> Building configuration... >> >> Current configuration : 107 bytes >> ! >> interface FastEthernet0/24 >> switchport mode dynamic desirable >> spanning-tree mst 0 port-priority 64 >> End >> >> But Cat 1 still shows the secondary link to Cat3 as its root port, and >> does not appear to have learned the preferred priority to Cat2. >> >> >> Cat1-MST(config-if)#do sh span >> >> MST0 >> Spanning tree enabled protocol mstp >> Root ID Priority 24576 >> Address 0018.baf8.5a80 >> Cost 0 >> Port 21 (FastEthernet0/21) >> Hello Time 3 sec Max Age 15 sec Forward Delay 15 sec >> >> Bridge ID Priority 32768 (priority 32768 sys-id-ext 0) >> Address 000b.be96.d800 >> Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec >> >> Interface Role Sts Cost Prio.Nbr Type >> ------------------- ---- --- --------- -------- >> -------------------------------- >> Fa0/1 Desg FWD 200000 128.1 P2p >> Fa0/4 Desg FWD 200000 128.4 P2p >> Fa0/5 Desg FWD 200000 128.5 P2p >> Fa0/11 Desg FWD 2000000 128.11 Shr >> Fa0/21 Root FWD 200000 128.21 P2p >> Fa0/23 Altn BLK 200000 128.23 P2p >> >> Interface Role Sts Cost Prio.Nbr Type >> ------------------- ---- --- --------- -------- >> -------------------------------- >> >> Fa0/24 Altn BLK 200000 128.24 P2p >> >> I have shut/no shut all relevant interfaces, but still no change. I have >> even removed and re-added the port-priority config on Cat2. >> >> Any ideas what I may be missing? >> >> Thanks, >> >> Rob >> _______________________________________________ >> For more information regarding industry leading CCIE Lab training, please >> visit www.ipexpert.com >> >> Are you a CCNP or CCIE and looking for a job? Check out >> www.PlatinumPlacement.com >> >> http://onlinestudylist.com/mailman/listinfo/ccie_rs >> > > > _______________________________________________ For more information regarding industry leading CCIE Lab training, please visit www.ipexpert.com Are you a CCNP or CCIE and looking for a job? Check out www.PlatinumPlacement.com http://onlinestudylist.com/mailman/listinfo/ccie_rs
