2

I have long distance connectivity between west coast to east cost datacenter via our ISP network as you can see that in following diagram (BSN=Boston, PHX=Phoenix)

enter image description here

Now when ISP-A or ISP-B doing any maintenance during that time PHX-SW1 switch putting interface in LOOP* (ofc because BPDU is missing during ISP is down but curious why its thinking its loop and trying to protect link? is there anything i should be doing for long distance STP tuning?)

PHX-SW1-config:

interface Ethernet1/1
  switchport mode trunk
  switchport trunk allowed vlan 4

This is the condition when ISP-A is down

Interface        Role Sts Cost      Prio.Nbr Type
---------------- ---- --- --------- -------- --------------------------------
Eth1/1           Root BKN*4         128.1    P2p *LOOP_Inc
Eth1/2           Root FWD 4         128.2    P2p 
Eth1/3           Altn BLK 4         128.3    P2p 
Eth1/4           Desg FWD 4         128.4    P2p 

PHX-SW1# 2021 Aug 31 21:02:10 PHX-SW1 %$ VDC-1 %$ %STP-2-LOOPGUARD_BLOCK: Loop guard blocking port Ethernet1/1 on VLAN0004.

when ISP-A is up it unblock interface and everything back to normal.

Update - 1

BSN-SW1 and BSN-SW2 is in vPC

BSN-SW1

BSN-SW1# show spanning-tree vlan 4

VLAN0004
  Spanning tree enabled protocol rstp
  Root ID    Priority    4100
             Address     0023.04ee.be01
             This bridge is the root
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

  Bridge ID  Priority    4100   (priority 4096 sys-id-ext 4)
             Address     0023.04ee.be01
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

Interface        Role Sts Cost      Prio.Nbr Type
---------------- ---- --- --------- -------- --------------------------------
Po999            Root FWD 4         128.5094 (vPC peer-link) Network P2p 
Eth1/1           Desg FWD 4         128.1    P2p 
Eth1/2           Desg FWD 4         128.2    P2p 

BSN-SW2

BSN-SW2# show spanning-tree vlan 4

VLAN0004
  Spanning tree enabled protocol rstp
  Root ID    Priority    4100
             Address     0023.04ee.be01
             This bridge is the root
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

  Bridge ID  Priority    4100   (priority 4096 sys-id-ext 4)
             Address     0023.04ee.be01
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

Interface        Role Sts Cost      Prio.Nbr Type
---------------- ---- --- --------- -------- --------------------------------
Po999            Desg FWD 4         128.5094 (vPC peer-link) Network P2p 
Eth1/1           Desg FWD 4         128.1    P2p 
Eth1/2           Desg FWD 4         128.2    P2p 

PHX-SW1

PHX-SW1# show spanning-tree vlan 4

VLAN0004
  Spanning tree enabled protocol rstp
  Root ID    Priority    4100
             Address     0023.04ee.be01
             Cost        4
             Port        1 (Ethernet1/1)
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

  Bridge ID  Priority    32772  (priority 32768 sys-id-ext 4)
             Address     5254.0018.7292
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

Interface        Role Sts Cost      Prio.Nbr Type
---------------- ---- --- --------- -------- --------------------------------
Eth1/1           Root FWD 4         128.1    P2p 
Eth1/2           Altn BLK 4         128.2    P2p 
Eth1/3           Altn BLK 4         128.3    P2p 
Eth1/4           Desg FWD 4         128.4    P2p 

PHX-SW2

PHX-SW2# show spanning-tree vlan 4

VLAN0004
  Spanning tree enabled protocol rstp
  Root ID    Priority    4100
             Address     0023.04ee.be01
             Cost        4
             Port        1 (Ethernet1/1)
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

  Bridge ID  Priority    32772  (priority 32768 sys-id-ext 4)
             Address     5254.0008.b08b
             Hello Time  2  sec  Max Age 20 sec  Forward Delay 15 sec

Interface        Role Sts Cost      Prio.Nbr Type
---------------- ---- --- --------- -------- --------------------------------
Eth1/1           Root FWD 4         128.1    P2p 
Eth1/2           Desg FWD 4         128.2    P2p 
Eth1/3           Desg FWD 4         128.3    P2p 
10
  • 2
    Why are you using STP in a meshing scenario that would work so much better using routed links?
    – Zac67
    Commented Aug 31, 2021 at 21:33
  • 3
    You need to set your root bridge. Don’t depend on luck to set it for you.
    – Ron Trunk
    Commented Aug 31, 2021 at 23:30
  • 1
    @RonTrunk I have my root bridge already set on one of switch, do you think root bridge placement causing here?
    – Satish
    Commented Sep 1, 2021 at 11:15
  • 1
    @Satish I think you need to post full sanitized configs for both switches involved, to start with. Maybe it’ll help if we can see the full picture.
    – Jesse P.
    Commented Sep 1, 2021 at 12:20
  • 2
    You might have a unidirectional link. The switch sees the Ubuntu server as its root -- that's a problem. Type show spanning-tree vlan 4 on all switches and add that info to your question.
    – Ron Trunk
    Commented Sep 1, 2021 at 13:00

1 Answer 1

2

xSTP is not a good choice when the link status (up/down) does not reflect actual L2 connectivity.

Your diagram shows an L2 connection between BSN-1 and PHX-1 across different ISPs. Since that is very unlikely, you seem to have some L2 tunneling going on there (VPN?). That tunneling possibly hides the actual link status when connectivity is lost, at least for some time.

You should consider switching to routed L3 links that don't require xSTP for removing bridge loops. Assuming you're using VPN between locations, OSPF between the routers should provide more reliable link status and route propagation.

3
  • I know L3 is best choice but at present i have no L3 switch on both end, until i go buy 5 digit license i am not able to convert them in routed link. (I am ok with loose connectivity because this is not impacting revenue so just looking for some interim solution where i can survive and later when we have money just replace gears.
    – Satish
    Commented Sep 1, 2021 at 15:44
  • @Satish: Maybe abuse LACP as poor man's link state protocol across that link? In extenso: configure LACP with fast timers on both ends, with a single member interface facing that chain of ISPs. This might (more quickly) help both ends understand if the link is up or not. Be careful though with possible situations for VPC with Orphan Ports Commented Sep 1, 2021 at 16:48
  • 1
    @Satish I was speculating about the routing equipment that you appear to have in place already - the problem should be solved there.
    – Zac67
    Commented Sep 1, 2021 at 17:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.