Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 425
  • Last Modified:

Simple frame relay question

A guy at work said are WAN connection is a full mesh Frame Relay.
However, when he showed me the equipment, only one serial interface on our router was occupied.

We have 4 sites. Wouldnt a full mesh require a separate T1 for every site (hence we would have more serial connections on our router?) .

In other words, does full mesh frame = a lot of serial devices connecting to your router?
  • 2
  • 2
2 Solutions
Not necessarily. A single T1 can have many individual PVC's with either a single point-to-multipoint interface, or multiple point-to-point subinterfaces..
for a 4-site full mesh, each site has at least 3 PVC's within the single T1




Not at all, and this is precisely the advantage frame relay has over point to point circuits which would require several individual circuits and interfaces to achieve the same thing. This is also why I think frame relay will be around for some time to come despite my belief that it is far from being a growth industry. It might not be state of art, but in certain circumstances it is still hard to beat, and this is one of those cases where frame relay is hard to beat.  
dissolvedAuthor Commented:
so multipoint frame is considered full mesh
point to point is partial

So in full mesh, do you need to have an interface on your router for each network you connect with?  Or are your different site's routers in the same subnet?  So site A's  router s0 is, site B's router so is  etc etc etc?

Multipoint frame is not necessarily full mesh.
point-point can be full mesh.

As in my post above, if A, B, C and D all have point-to-multipoint interfaces, and we assign each one an IP address in the same subnet, then we use map statements to map a destination IP to a DLCI. The prefered method is to use multiple point-to-point sub-interfaces, with each link in its own subnet, then use dynamic routing protocol across the links.
One advantage to using subinterfaces, is that if a PVC drops out, the sub-interface drops and can generate a SNMP trap, using "interface down" messages can be a trigger event. Using point-multipoint and IP mappings, there is no trigger event if a PVC drops, but the interface stays up.

The difference between partial mesh and full mesh using the example I posted above, as long as all 4 sites maintain one link to each of the other 3 sites, this is a full mesh. If, for example D only had a PVC to A, then to get from D<->C, you must pass through A, while B<->C is a direct path withough going through A..

This document has some good graphics showing the differences between star/mesh/partial mesh/hybrid topologies

The thing is with frame relay you only need one physical interface, and through the use of DLC’s you create multiple virtual interfaces. So you have one T1 coming in per location, through which you connect to multiple locations using the multiple virtual circuits that frame relay allows.

In away you can kind of think of frame relay network as functioning like an Ethernet switch when you have a full mesh network configured. Any connection to it can talk to any other connection to it. Unlike when you have a point-to-point network where you need a separate link from every location to every other location you want to communicate to.  


Featured Post

Concerto Cloud for Software Providers & ISVs

Can Concerto Cloud Services help you focus on evolving your application offerings, while delivering the best cloud experience to your customers? From DevOps to revenue models and customer support, the answer is yes!

Learn how Concerto can help you.

  • 2
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now