JC Dill wrote: > I'm in the process of spec'ing out a meshed Frame Relay network and would > like to have as much physical diversity in the mesh as possible. The plan > is to have 2 or more FR lines into each location, running in separate > bundles, preferably in separate conduits and thru separate paths back to > the CO, and then have each FR line terminated on a different switch. > I'd love suggestions on how to actually achieve this. Are there any > special terms to use to specify this? Is there any way to determine if we > are actually *getting* it, if it's so specified? Are you completely within one RBOC's coverage area, or are you going inter-LATA? If the former, you'll just need to explain what you want to their enineers doing the build, and make sure they agree to it before you place the order. Most places, you'll be able to get platform diversity as well. That is, one set of connections on Cascade BSTDXes and the other set on Newbridge 36/170s, since most shops have fairly complete coverage with both. They obviously have different bugs and peculiarities, partiicularly when it comes to their NMSes, so diversity there will help you at least as much as diversity in the local loop. If you're going inter-LATA, it's pretty simple... You'll just need to pick two LD companies. Since the LD companies each have their own NNI, and those NNIs are typically fed off of different switches on the RBOC side, there's no overlap except what you may get from two LD companies sharing a fiber somewhere. That you can probably take care of through sensible routing and a little knowledge of the fiber infrastructure. For instance, leave DC southbound on Sprint and northbound on Worldcom, and you'd be less likely to share a bundle than if you had two smaller carriers, both taking you from DC to NYC. The diverse local loops will be your least tractable problem. -Bill