RSTP Blocking Port

[Gurinder_1]

I have problems setting up HP 2900 switch to a Foundry RX16. What i am trying to achieve is that the RX16 is our core switch and the HP2900 switch is in a remote location. There is a 10GB link connecting them with 1GB link as a failover. The 1GB link isn't connected yet. On the Rx16 i have enabled rstp, we have 2 vlans (101 and 102). The problem is that when i have enabled rstp on the HP switch on the other end, using the following commands:

spanning
spanning-tree config-name "remote"
spanning-tree force-version RSTP-operation

The running-config is:

vlan 101
name "MANAGEMENT"
untagged 1
tagged 24,A1
no ip address
exit
vlan 102
name "MPCPRODSOHO"
untagged 2-9,15-17,A2
ip address 172.16.1.228 255.255.0.0
tagged 24,A1
exit

The 10GB uplink port gets blocked, doing a 'show span' shows:

Multiple Spanning Tree (MST) Information

STP Enabled : Yes
Force Version : RSTP-operation
IST Mapped VLANs : 1-4094

Filtered Ports :
Switch MAC Address : 001c2e-152b00
Switch Priority : 32768
Max Age : 20
Max Hops : 20
Forward Delay : 15

CST Root MAC Address : 0012f2-3f3b00
CST Root Priority : 32768
CST Root Path Cost : 2000
CST Root Port : 1

IST Regional Root MAC Address : 0019bb-201980
IST Regional Root Priority : 32768
IST Regional Root Path Cost : 0
IST Remaining Hops : 20

Port Type | Cost rity State | Bridge Time PtP Edge
---- --------- + --------- ----- ---------- + ------------- ----- --- ----
1 100/1000T | Auto 128 Disabled | 0012f2-3f3b00 2 Yes Yes
2 100/1000T | Auto 128 Disabled | 0012f2-3f3b00 2 Yes Yes
3 100/1000T | Auto 128 Disabled | 0012f2-3f3b00 2 Yes Yes
4 100/1000T | Auto 128 Disabled | 0012f2-3f3b00 2 Yes Yes
5 100/1000T | Auto 128 Disabled | 0012f2-3f3b00 2 Yes Yes
A1 10GbE-SR | 2000 128 Blocking |
A2 10GbE-CX4 | 2000 128 Forwarding | 0019bb-201980 2 Yes Yes

Port A1 is tagged on vlan 101 and 102. But if i unplug port 1, which is untagged in vlan 101. Then the main uplink port gets back to a forwarding state. This switch is running 'T.13.45' firmware.

As a test i replicated the remote switch locally and installed the latest firmware (T.13.63) and when i run spanning-tree on this switch by running the same commands, port A1 was forwarding, whilst all the other ports were blocked.

The output of 'show span' of this switch is:

Multiple Spanning Tree (MST) Information

STP Enabled : Yes
Force Version : RSTP-operation
IST Mapped VLANs : 1-4094
Switch MAC Address : 0019bb-201980
Switch Priority : 32768
Max Age : 20
Max Hops : 20
Forward Delay : 15

Topology Change Count : 1
Time Since Last Change : 29 secs

CST Root MAC Address : 0012f2-3f3b00
CST Root Priority : 32768
CST Root Path Cost : 2000
CST Root Port : A1

IST Regional Root MAC Address : 0019bb-201980
IST Regional Root Priority : 32768
IST Regional Root Path Cost : 0
IST Remaining Hops : 20

Root Guard Ports :
TCN Guard Ports :
BPDU Protected Ports :
BPDU Filtered Ports :
PVST Protected Ports :
PVST Filtered Ports :

| Prio | Designated Hello
Port Type | Cost rity State | Bridge Time PtP Edge
----- --------- + --------- ----- ---------- + ------------- ----- --- ----
1 100/1000T | 20000 128 Blocking | 0019bb-201980 2 Yes No
2 100/1000T | 20000 128 Blocking | 0019bb-201980 2 Yes No
3 100/1000T | 20000 128 Blocking | 0019bb-201980 2 Yes No
4 100/1000T | 20000 128 Blocking | 0019bb-201980 2 Yes No
A1 10GbE-SR | 2000 128 Forwarding | 0012f2-3f3b00 2 Yes No
A2 10GbE-CX4 | 2000 128 Blocking | 0019bb-201980 2 Yes No

Not sure why this is happening. No other switch on our network is running 'rstp' or any other form of spanning tree.

Not sure what to do, any help would be greatly appreciated.

Thanks.

[Pieter 't Hart]

>>>
But if i unplug port 1, which is untagged in vlan 101. Then the main uplink port gets back to a forwarding state
<<<
combined with
>>>
CST Root MAC Address : 0012f2-3f3b00
CST Root Priority : 32768
CST Root Path Cost : 2000
CST Root Port : 1
<<<

shows that a switch reachable through port-1 is selected as CST root.
>>>
No other switch on our network is running 'rstp' or any other form of spanning tree.
<<<
you need to investigate the config of the switch at port-1, or any other device connected to this switch. (rogue switch?)

in your test environment you don't have a switch with the same characteristics connected to this port-1 to select it as root.

here the switch is it's own "IST regional root :
Switch MAC Address : 0019bb-201980
IST Regional Root MAC Address : 0019bb-201980
while in production it's not :
Switch MAC Address : 001c2e-152b00
IST Regional Root MAC Address : 0019bb-201980

As you run only a single MSTP instance, the same stp-topology counts for all vlan's

[Pieter 't Hart]

sorry, my last remark :
>>>
As you run only a single MSTP instance, the same stp-topology counts for all vlan's
<<<
is not specific enough.

Because of the foundry network in between, that probably isn't configured for MSTP, your topology is split into MSTP regions.

a remote MSTP region, the foundry STP region and your local MSTP region.
each maintain their own "internal" (IST) topology.
in adition there is an "overall" topology
(CST)
look at image from advanced trafic management guide

[Gurinder_1]

Thanks for your reply.

When you mentioned:

--
"to investigate the config of the switch at port-1, or any other device connected to this switch. (rogue switch?)"
--

The rogue switch in this instance is a HP2900-48 port switch that isn't running anything clever like MSTP or RSTP, the running-config is:

vlan 1
name "DEFAULT_VLAN"
untagged A1-A4
no untagged 1-48
no ip address
ip igmp
exit
vlan 101
name "MANAGEMENT"
untagged 1-48
ip address 10.64.1.184 255.255.0.0
exit
sntp server priority 1 172.16.3.6
no ip ssh
password manager
password operator

--
"in your test environment you don't have a switch with the same characteristics connected to this port-1 to select it as root. "
--

Whilst the other switch that is in the test environment on port-1 is a HP2648 and that again isn't running anything to clever:

vlan 1
name "DEFAULT_VLAN"
untagged 41-49
no ip address
no untagged 1-40,50
exit
vlan 101
name "MANAGEMENT"
untagged 1-40,50
ip address 10.64.1.159 255.255.0.0
exit

So i don't quite understand how port-1 became root. I have a downtime schedule booked soon, so on the rx16 i am going to run the following:

rstp single
rstp single priority 10
rstp single ethernet x/y admin-pt2pt-mac (run this on all the uplink ports)

This should hopefully make the RX16 the root bridge. This should stop the HP switches from try to make other ports that are connected into rogue switches the root. I will also read the image from advanced traffic management guide too.

Does this make sense ? can anyone see a problem that this might cause ?

Thanks.

[Pieter 't Hart]

Hi there

If the 2900 is not configured for spanning-tree, then some other device at this side of the network is!
So with rogue i mean that someone connected it's own switch to one of your network ports of the 2900.
Try "show mac-address 0012f2-3f3b00" at this 2900 to locate the port where the device with this mac-address is connected.

[Gurinder_1]

The switch with the MAC address of 0012f2-3f3b00 is the Foundry Rx16. This is rstp at the moment.

[Pieter 't Hart]

so it's also connected through port-1
not only A1 or 24

CST Root MAC Address : 0012f2-3f3b00
CST Root Priority : 32768
CST Root Path Cost : 2000
CST Root Port : 1

check cabling on the 2900 so it doesn't go back to the Rx16

[Gurinder_1]

Sorry i think i have confused matters slighty. As A1 should be the root port as that is the only link to the RX16 (the core switch), this is a 10Gb link. Port-1 is connecting to another switch (HP) that isn't running RSTP at all.

[Richard Brodie_1]

"As A1 should be the root port as that is the only link to the RX16 (the core switch)."

Presumably, this assumption is wrong. It's hard to tell without a diagram.

And what on earth are you doing with the test system that all ports except one are blocked?

[Gurinder_1]

"Presumably, this assumption is wrong. It's hard to tell without a diagram."

Port A1 should be the root, since this is a 10GB link whilst all the other ports are only 1Gb. At the moment every switch we have hangs of the Rx16. Port A1 (uplink) connects into the RX16, once RSTP is running correctly between these switches, i will make port 24 (1GB port) into the backup link to the RX16.

"And what on earth are you doing with the test system that all ports except one are blocked?"

I am not sure how this happened, as i did was enable rstp.

If it helps i could do a diagram. Thanks.