Is it true MS switches don't support some form of MSTP?

SOLVED
DCrab
Here to help

Is it true MS switches don't support some form of MSTP?

This can't be right.. do Meraki switches not support a spanning tree per vlan? I can't imagine turning STP on, if this were the case, in a multiple VLAN.. am I missing something?

1 ACCEPTED SOLUTION

Most customers really don't need spatial re-use of links, for the extra bandwidth - or can achieve spacial reuse through LACP between stacks.   So if customers don't really need something, why make a product more complex by including it?

 

While MX may not participate in (R)STP, it happily passes BPDUs, allowing you to build topologie with loops, for resilience.   The control of the loops is based upon your STP root bridge, in the usual way.

 

If Cisco customers do need these extra facilities, Catalyst will work well for them.

View solution in original post

11 REPLIES 11
UCcert
Kind of a big deal

Hi @DCrab 

 

Meraki switches run RSTP and not PVSTP or MSTP

 

https://documentation.meraki.com/MS/Port_and_VLAN_Configuration/Configuring_Spanning_Tree_on_Meraki_...)

Darren O'Connor | uccert.co.uk
https://www.linkedin.com/in/darrenoconnor/

I'm not an employee of Cisco/Meraki. My posts are based on Meraki best practice and what has worked for me in the field.

wow.. am surprised by this.. don't think I've used any managed switch in last 10 years that didn't support MSTP or PVSTP.. wondered why I was getting all kinds of STP traffic flipping ports status' around.. I've got a pretty simple network.. might just turn RSTP off.. 

RaphaelL
Building a reputation

@DCrab  Well no surprise here from Meraki.  The lack of support of industry standards in pretty common. Also MX do not run STP so here goes another surprise.

Most customers really don't need spatial re-use of links, for the extra bandwidth - or can achieve spacial reuse through LACP between stacks.   So if customers don't really need something, why make a product more complex by including it?

 

While MX may not participate in (R)STP, it happily passes BPDUs, allowing you to build topologie with loops, for resilience.   The control of the loops is based upon your STP root bridge, in the usual way.

 

If Cisco customers do need these extra facilities, Catalyst will work well for them.

View solution in original post

PhilipDAth
Kind of a big deal

>The lack of support of industry standards in pretty common

 

Neither spanning tree (802.1d) nor RSTP (802.1w) have any concept of VLANs.  Certainly, neither standard is "per-VLAN".  Both standards require the packets to be transmitted without tags.

 

Some vendors, like Cisco Enterprise, chose not to follow the standard and changed it to transmit the packets on all VLANs.  I believe Cisco Enterprise did this to improve load balancing - but this problem no longer exists now that we have channelling technology (and have done for many decades now).  So it is solving a problem that no longer exists, but they can't go back now to following the standard without causing major breakage.

 

Meraki came into existence after this problem was solved.  There is no point in getting them to also break the standard to solve a problem that does not exist.

 

MST is multi-instance, not per VLAN.  If you map an instance per VLAN you'll get this impact, but I don't know of any site that actually takes it to this level.  You might put downstream buildings or groups into an instance, but not per VLAN.

I’ve mostly worked with Cisco switches in the past 10 years, so pretty used to PVST. And I’ve fed switches from VLAN access lines.. I did that on Meraki stuff and they really don’t like that.. So as suggested, I should be using trunks to each switch with the native VLAN a member and that might clear up the STP ‘unstableness’. I’ve only got 50 active ports between 2 switches so a pretty simple network. I have lots of bandwidth and could use LACP (assuming MX75 supports it) if I needed anything additional. So my issue seems to be my network design rather than lack of feature on the switches.

RaphaelL
Building a reputation

MXs do not support LACP or any form of LAG.

> Don’t support LAG/LACP

 

Grin.. well it is a pretty small network.. so performance is not an issue.. one trunk line per switch should be fine. 

KarstenI
Kind of a big deal

Meraki first goal is simplicity. It doesn't always work as expected but if you use the system as intended you get an easily manageable network that really saves you time compared to the traditional Cisco gear..

KarstenI
Kind of a big deal

As @UCcert said, that's what it is. And with x*10GBit between Access and Core I actually don't care any more ... 😉

There is also one exception to this, the MS390 uses MST, only single instance but MST ... 😉

redsector
Head in the Cloud

We are running a Cisco Classic network with Cisco Catalyst switches. They are running with PVST. But we connected Meraki-Switches without any issues. You need to have VLAN1 as standard VLAN. And don´t connect Cisco classic, Meraki, and again Cisco Classic in a row. In the end our core and some distri-switches are Cisco classic and userswitches are Meraki.

 

https://documentation.meraki.com/MS/Deployment_Guides/Advanced_MS_Setup_Guide

Get notified when there are additional replies to this discussion.
Welcome to the Meraki Community!
To start contributing, simply sign in with your Cisco account. If you don't yet have a Cisco account, you can sign up.
Labels