Connecting Multiple MST Regions.

[robert.gillen]

Hello this is my first post so excuse me and my knowledge on MST regions is going round and round in circles.

I Currently have 13 switches in partial mesh. I am on the boundary of these devices being in the standard STP "diameter". Due to the variability of  the connection I do not want to play around with timers etc. The Link is pretty critical so i dont want to take chances and the Links I have available and to which switches they go are fixed.

Anyway long story short.

If i devide the switching domain into.

############                                                         

# REGION 0 #    Partial Mesh

############                                                         

|                    |

|                    |

|                    |

Boundary       Boundary

Switch 1         Switch 2

|                    |

|                    |

|                    |

|                    |

|                    |

############                                                         

# REGION 1 #      Partial Mesh

############   

Will CIST Block off the link to from region 1 -> boundary 2 -> region 0.

and allow diameter of switches to be managed by each MST region??? or would the Boundary switches need to be apart of a specific region.

or possibly implement a layer 3 boundary where the boundary switches are.

This is a fairly odd STP operation design for me so i could be completely wrong in my esign and thoughts.

In advance thanks for any feedback and help given .

Regards Robert.

[Peter Paluch · Accepted solution]

Hello Robert,

I do not quite understand what you are trying to accomplish by splitting your network into several MST regions and your requirement about region diameter. You write:

I Currently have 13 switches in partial mesh. I am on the boundary of  these devices being in the standard STP "diameter". Due to the  variability of  the connection I do not want to play around with timers  etc

First of all, the notion of timers is irrelevant in MSTP as it is not driven by timers; rather, it behaves like a multi-instance RSTP (which it really is). Second, there is a notion of diameter in MSTP but it is just a hop counter for sanity checks - it is not directly influencing the convergence time. Within a single region, you can always manipulate the maximum region diameter using the spanning-tree mst max-hops command with the valid value in the range 1-255. Your current diameter of roughly 13 is not a problem. The default setting of this command is 20.

Best regards,

Peter

[robert.gillen]

Thanks Peter you have essentially answered my question anyway. I was still a bit confused on the RTSP convergence.

Cheers robert

Sent from Cisco Technical Support iPhone App