isis mesh-group

[sachin30720041]

Hello All,

I am trying to understand the isis mesh-group.

On R1, g1,g2 and gi3 are part of the same mesh-group. Now If a new loopback(44.44.44.44/32) is added on R4 and advertised in ISIS then R1 will receive the new LSP on G1and G2 and R1 should not flood the LSP through G3. As a result, R5 should not have 44.44.44.44/32. Am I correct in my understanding?

R5

|

(gi3)

R1(g2)-----R2

(gi1)              |

|.                   |

|.                   |

R3---------R4

[Flavio Miranda]

Should not due the rule:

" LSPs that are first received on subinterfaces that are part of a mesh
group are flooded to all interfaces except those in the same mesh group."

Besides, Mesh group is meant to highly meshed tolopogy which is not the case for R5.

[sachin30720041]

The Mesh group is configured on R1. Gi1, Gi2, and Gi3 are under the same mesh group. Any new LSP arriving at Gi1 or Gi2 should not be flooded to Gi3. Correct?

[Flavio Miranda]

That´s why understand from the above statement, right;

[sachin30720041]

You mean to say if there are multiple links between R1 and R5. And all the interfaces on R1 pointing to R5 are part of the same mesh group. Then only one interface will forward the LSP towards the R5. The rest of the interfaces will block the new LSP. Am I correct?

[Flavio Miranda]

"When link-state PDUs are being flooded throughout an area, each router within a mesh group receives only a single copy of a link-state PDU instead of receiving one copy from each neighbor, thus minimizing the overhead associated with the flooding of link-state PDUs"

[Flavio Miranda]

But you may know that this is quite rudimentar and can fail in case the design was not done correct.

 Even on the RFC they estate:

 "<span;>The concept of using numbered mesh groups also suffers from the complexity and reliance on static configuration, making the topologies brittle. Loosing a transit link can partition LSP flooding in unpredictable ways, requiring the periodic flooding of CSNPs to synchronize databases. In large networks, CSNPs become large and also consume bandwidth."