cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
448
Views
3
Helpful
13
Replies

STP/VLAN Redundancy Question

NetworkNewbie37
Level 1
Level 1

Hello everyone,

     I'm brand new to posting on this forum, and I hope I'm not submitting this in the incorrect location. Please let me know if there is a proper location for discussion such as this to be posted to, I appreciate the understanding.

     A question I have for the experts, is it common place for companies to separate root bridges per different VLANs? I'm starting to see the need of features such as Root Guard, BPDUGuard, Loop Guard, etc. to ensure the primary root bridge is not "compromised" and to maintain a stable topology. 

      For example; is there a major need in the field to ensure VLAN 5 is preserved on SwitchA as the root primary, while VLAN 12 maintains on SwitchB as it's own root primary? I understand it is used for redundancy purposes, but receiving insight from the people who work with this directly would be very beneficial for my own understanding!

Again I hope I'm not posting in the wrong location, and I do appreciate any and all feedback I may receive.

Thank you for your time.

1 Accepted Solution

Accepted Solutions

 

  - @NetworkNewbie37       A solid core networking infrastructure will always remain up by itself.
                                           For selecting cisco products look at : https://www.cisco.com/c/en/us/products/switches/switch-selector.html
                                          and select the offered solutions according to your needs

  M.



-- Let everything happen to you  
       Beauty and terror
      Just keep going    
       No feeling is final
Reiner Maria Rilke (1899)

View solution in original post

13 Replies 13

Mark Elsen
Hall of Fame
Hall of Fame

 

  - @NetworkNewbie37  It's  not  common to separate root bridges per VLANS, actually it's the reverse where a central core becomes the root bridge for the whole network, taking into account redundancy features for core networking ,

   M.



-- Let everything happen to you  
       Beauty and terror
      Just keep going    
       No feeling is final
Reiner Maria Rilke (1899)

Hello Mark, and thank you for taking the time to reply.

If I understand correctly, it's often set to the core root bridge to be the heart for a network/switch topology, while using tools to maintain that core in an uptime state?

 

  - @NetworkNewbie37       A solid core networking infrastructure will always remain up by itself.
                                           For selecting cisco products look at : https://www.cisco.com/c/en/us/products/switches/switch-selector.html
                                          and select the offered solutions according to your needs

  M.



-- Let everything happen to you  
       Beauty and terror
      Just keep going    
       No feeling is final
Reiner Maria Rilke (1899)

Understood, thank you for the insight Mark.

I hope you have a great day!

Joseph W. Doherty
Hall of Fame
Hall of Fame

Your question is in the correct forum.


@NetworkNewbie37 wrote:

      For example; is there a major need in the field to ensure VLAN 5 is preserved on SwitchA as the root primary, while VLAN 12 maintains on SwitchB as it's own root primary? I understand it is used for redundancy purposes, but receiving insight from the people who work with this directly would be very beneficial for my own understanding!


A major need?  That's an it depends answer.

Different roots, per VLAN (within PVST), is not for redundancy, it (and PVST) is for optimizing performance.

By having different VLAN roots, I can have different L2 topologies on the same physical topology, which may allow me to spread out switch and link loading.

On modern networks, where typical modern switches are line-rate for all ports, and using Etherchannel, the need for PVST, for optimizing performance, is somewhat uncommon, and if you need to do it at all, you can now use MST, and avoid being limited to using proprietary PVST.  Further, with the advent of "inexpensive" L3 switches, STP can pretty much be ignored and used mainly to preclude accidental L2 loops.

Hi Joseph,

Thank you for the response, I think I understand.

The use of a feature such as STP/PVST (or RSTP) would less likely be used for redundancy, and more so for preventing L2 loops as a whole. Which I appreciate you mentioning EtherChannel, because I just began learning about EtherChannel today!

Originally, STP was for designed redundancy.  I.e. a L2 design could have warm backup links.

Cisco's PVST allowed redundant L2 links to carry traffic, provided you had multiple VLANs.

Etherchannel (e.g. LACP) also allows multiple L2 links to carry traffic, but not limited by VLANs, but usually limited to connections between same two devices.

L3 also supports multiple paths, usually allowing active usage of those paths but also isn't limited by VLANs.  With L3 topologies, often no need for STP, by design, but as someone can accidentally create a L2 loop, STP normally used to preclude that.

Thank you for the insight Joseph,

At this point it seems Etherchannel by default is a L2 feature, however, just as you can with a standard trunk port, and as you stated it's not limited by VLANs

Would this mean that Etherchannel can be combined as a L2/L3 feature, or is it generally referred to as a L2 feature?


@NetworkNewbie37 wrote:

Thank you for the insight Joseph,

At this point it seems Etherchannel by default is a L2 feature, however, just as you can with a standard trunk port, and as you stated it's not limited by VLANs

Would this mean that Etherchannel can be combined as a L2/L3 feature, or is it generally referred to as a L2 feature?


Etherchannel, bundles multiple physical links into one logical link.  Like other links, it can be used for L2 or L3, the latter, L3, of course, running over L2.

If I set up an Etherchannel between a pair of switches, the port-channel interface, could be, I believe, configured as an access port, although more likely, for L2, to be configured as a trunk port.  The port-channel interface, could also be configured as a "routed" port, i.e. have an IP addresses.  It also, like other ports, might be configured as L2 port, but with a dedicated VLAN running across it, supporting L3 routing concurrently with other L2 traffic running across it concurrently.

BTW, @NetworkNewbie37 do you have a copy of Cisco's Packet Tracer?  If so, you can experiment with a lot of basic features.  However, if using PT, do be aware, for the features it supports, they aren't always faithful to real platforms.

Cisco, I believe, now provides a 5 node "free". version of CML  It appears to be very faithful to real platforms, including, it appears, all features, but it's much more resource intensive, and I don't believe it supports any low end switches.

The two products, though, can make a nice pair for learning.

I have been working with Packet Tracer, it seems to be a necessity when trying to learn all of this. Without any hands on, I don't think I could understand any of the topics I've covered so far.

I was unaware of the 5 node CML, I will look into that for sure. Also as you stated it seems, depending on the need, Etherchannel can either be L2 and/or L3 at the same time.

Thank you again for the feedback!

depending on the need, Etherchannel can either be L2 and/or L3 at the same time.

Again, that's not specific to Etherchannel.  Also again, Etherchannel bundles multiple links to form one logical link.  That logical link's interface, more-or-less, has the same options as do other interfaces.

Also, Etherchannel isn't limited to switches.  Cisco routers often support it as do some servers.

Etherchannel serves two purposes, more aggregate bandwidth, and link redundancy.

Thank you for taking the time in responding to my questions.

I hope you have a great day Joseph!