11-09-2013 08:44 PM - edited 03-07-2019 04:30 PM
I have two switches connected with two cables. One port is configrued as trunk and another port is configured as access on each end of the switch respectively. Spanning tree has blocked the port configured in trunk mode, due on higher port id. I know I am not supposed to configure the port connecting to switch in access mode. I am tyring to understand, does spanning tree protocol work independent of modes configured? And, I think yes.
Thanks in advance!
Solved! Go to Solution.
11-10-2013 12:17 AM
Hi,
I think you are right, STP is independent of mode configure.
You can verify this by configuring higher port id as access port and lower as trunk.
Thanks
Shashi
11-10-2013 02:08 AM
I understand the scenario in the original question like this: Two swiches are interconnected to each other by one connection, which is configured as a trunk on both sides and also by another connection which is configured as access-ports on both sides. Please correct me if I'm wrong.
The original IEEE spanning-tree versions build a common spanning tree on a per-port basis, completely independent of the switchport modes.
However, the Cisco versions PVSTP+ and Rapid-PVSTP+ build their spanning-tree topologies on a per-VLAN basis (additional, special BPDUs are transmitted for this purpose). So the VLAN which is configured as the access-vlan on the non-trunk connection may result in blocking-state on the trunk-port, while other VLANs on the same trunk-port are in forwarding-state.
Of course, if you only have one VLAN (the default VLAN 1), it doesn't make a difference.
HTH
Rolf
11-11-2013 05:58 AM
Frames are tagged on egress on trunk ports. When frames are received ingress on trunk ports the tag is removed.
Daniel Dib
CCIE #37149
11-10-2013 12:17 AM
Hi,
I think you are right, STP is independent of mode configure.
You can verify this by configuring higher port id as access port and lower as trunk.
Thanks
Shashi
11-10-2013 12:48 AM
You have inconsistent port types. This can be detected by receiving tagged BPDUs on an access ports. This can also occur in the case of mismatched native VLANs. STP will block the offending VLAN(s) until the inconsistency has been cleared.
Daniel Dib
CCIE #37149
11-10-2013 02:08 AM
I understand the scenario in the original question like this: Two swiches are interconnected to each other by one connection, which is configured as a trunk on both sides and also by another connection which is configured as access-ports on both sides. Please correct me if I'm wrong.
The original IEEE spanning-tree versions build a common spanning tree on a per-port basis, completely independent of the switchport modes.
However, the Cisco versions PVSTP+ and Rapid-PVSTP+ build their spanning-tree topologies on a per-VLAN basis (additional, special BPDUs are transmitted for this purpose). So the VLAN which is configured as the access-vlan on the non-trunk connection may result in blocking-state on the trunk-port, while other VLANs on the same trunk-port are in forwarding-state.
Of course, if you only have one VLAN (the default VLAN 1), it doesn't make a difference.
HTH
Rolf
11-10-2013 03:29 AM
Hello Daniel,
I understand. What if I use standard STP VLAN 1, BPDU's will be sent untagged right?
Thanks,
Mugilan
11-10-2013 09:13 AM
Hi Mugilan,
I hope you don't worry if I try to answer the question.
On the access-ports, Cisco switches sends untagged IEEE BPDUs.
On a trunk, they also send those untagged IEEE BPDUs in VLAN 1; plus untagged, so called SSTP-BPDUs (Cisco proprietary), on the native VLAN (which is by default VLAN 1); and, if other VLANs exist, tagged SSTP-BPDUs on every non-native VLAN.
Sent from Cisco Technical Support Android App
11-11-2013 05:21 AM
Hi Fischer,
Absolutely, no problem. That was pretty clear. Thank you so much, One more question if you don't mind, If you configure a port, for example fa0/2, to be a part of vlan 2, when does the switch exactly tag the frames when it is ingress on fa0/2 or before it switches or floods the frame out?
Thanks,
Mugilan
11-11-2013 05:58 AM
Frames are tagged on egress on trunk ports. When frames are received ingress on trunk ports the tag is removed.
Daniel Dib
CCIE #37149
11-11-2013 09:06 AM
That makes perfect sense, Daniel. Thanks! Not sure why I thought that way
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide