Comware

I got 2 x A5500 switches and 1 x V1810 switch. On V1810 switch, I tried to create 2 redundant links to the respective A5500 switch but I got looping error.

 

Here's my setup.

 

On Core 1 and 2, port 32 are linked together and I had set them to be hybrid ports, all vlans tagged.

 

RSTP is enabled on both switches. I had set Core 1 as the primary root and Core 2 as secondary root. The rest STP values are default.

 

On Core 1, Gi1/0/1 is the fiber uplink to V1810 port 52.

 

On Core 2, Gi1/0/28 is the RJ45 uplink to V1810 port 48.

 

If I plug in fiber uplink from V1810 to Core 1, everything working fine.

 

With fiber uplink in place, I plug in RJ45, STP looping error occured. I can see on V1810 switch LEDs all blinking at the same time.

 

When I type dis stp bri on both A5500 switches, here's the output.

 

-----------------------------------------------------

[Core 1]dis stp bri
 MSTID      Port                         Role  STP State     Protection
   0        GigabitEthernet1/0/1         DESI  FORWARDING    NONE
   0        GigabitEthernet1/0/32        DESI  FORWARDING    NONE

 

 

[Core 2]dis stp bri
 MSTID      Port                         Role  STP State     Protection
   0        GigabitEthernet1/0/28         DESI  FORWARDING    NONE
   0        GigabitEthernet1/0/32        DESI  FORWARDING    NONE

 

 

 

 

[Core 1]dis stp
-------[CIST Global Info][Mode RSTP]-------
CIST Bridge         :0.3ce5-a667-d4a6
Bridge Times        :Hello 2s MaxAge 20s FwDly 15s MaxHop 20
CIST Root/ERPC      :0.3ce5-a667-d4a6 / 0
CIST RegRoot/IRPC   :0.3ce5-a667-d4a6 / 0
CIST RootPortId     :0.0
BPDU-Protection     :disabled
Bridge Config-             
Digest-Snooping     :disabled
CIST Root Type      :PRIMARY root
TC or TCN received  :74
Time since last TC  :0 days 0h:10m:1s

----[Port1(GigabitEthernet1/0/1)][FORWARDING]----
 Port Protocol       :enabled
 Port Role           :CIST Designated Port
 Port Priority       :128
 Port Cost(Dot1T)    :Config=auto / Active=20000
 Desg. Bridge/Port   :0.3ce5-a667-d4a6 / 128.1
 Port Edged          :Config=disabled / Active=disabled
 Point-to-point      :Config=auto / Active=true
 Transmit Limit      :10 packets/hello-time
 Protection Type     :None
  ---- More ----                 MST BPDU Format     :Config=auto / Active=legacy
 Port Config-             
 Digest-Snooping     :disabled
 Rapid transition    :false
 Num of Vlans Mapped :4
 PortTimes           :Hello 2s MaxAge 20s FwDly 15s MsgAge 0s RemHop 20
 BPDU Sent           :313             
          TCN: 0, Config: 0, RST: 313, MST: 0
 BPDU Received       :0             
          TCN: 0, Config: 0, RST: 0, MST: 0

 

[Core 2]dis stp
-------[CIST Global Info][Mode RSTP]-------
CIST Bridge         :4096.b8af-670a-ca80
Bridge Times        :Hello 2s MaxAge 20s FwDly 15s MaxHop 20
CIST Root/ERPC      :0.3ce5-a667-d4a6 / 20
CIST RegRoot/IRPC   :4096.b8af-670a-ca80 / 0
CIST RootPortId     :128.26
BPDU-Protection     :disabled
Bridge Config-
Digest-Snooping     :disabled
CIST Root Type      :SECONDARY root
TC or TCN received  :38
Time since last TC  :0 days 0h:12m:6s

 

----[Port28(GigabitEthernet1/0/28)][FORWARDING]----
 Port Protocol       :enabled
 Port Role           :CIST Designated Port
 Port Priority       :128
 Port Cost(Legacy)   :Config=auto / Active=20
 Desg. Bridge/Port   :4096.b8af-670a-ca80 / 128.28
 Port Edged          :Config=disabled / Active=disabled
 Point-to-point      :Config=auto / Active=true
 Transmit Limit      :10 packets/hello-time
 Protection Type     :Loop
 MST BPDU Format     :Config=auto / Active=legacy
 Port Config-
 Digest-Snooping     :disabled
 Num of Vlans Mapped :4
 PortTimes           :Hello 2s MaxAge 20s FwDly 15s MsgAge 1s RemHop 20
 BPDU Sent           :853
          TCN: 0, Config: 0, RST: 853, MST: 0
 BPDU Received       :0
          TCN: 0, Config: 0, RST: 0, MST: 0

--------------------------------------------------

 

I am puzzled. Shouldn't Core 2 Gi 1/0/28state be "DISCARDING" ?

 

Anyone can advise where could be the problem?

 

 

 

 

is STP configured on the V1810? If not and it is dropping bpdus it will loop anyway.

V1810 do not support STP. So based on your reply, that meant I cannot have redundant uplinks on the V1810 switch?

Not if the v1810 discards STP BPDUs (which actually seems to be tha case if you are able to loop it). One thing I came to think of was to connect the two 5500s in an IRF and use the distributed portchannel to dual connect the v1810.. but if it doesnt support STP it probably wont support portchannel either?

Hi Jason,

I was going to post similar question. Only difference is we have 2x 2910al-24G and 5x1810. Have you tried loop prevention mechanism either on 2910/1820 or both? I think the best it will come up with is permanently disabling redundant links so not much failover there as you would need to -re-enable them manually. It does allow you to have the interface re-enable itself after a certain number of seconds (min is 5s I think). It could perhaps cause intermittent issues as it detects loop and then shuts down the interface over and over again...

If that does not work the best we can come up is 2Gb LACP trunk from one 2910 to each 1810. Should 2910 fail customer wil have to manually switch cables over.

Any ideas?

Loop prevention mechanism on 1810 switch doesn't help. When loop occur, it doesn't shutdown at all.

 

I even tried loop-detection action command on the trunk ports on A5500. Doesn't help too.

Didn't u try USE Link aggregation (LACP ) between 1810 and 5500