Do people ever configure routers as Frame Relay switches in a production environment?

CodeBlox

In the lab, we have to use a router as a frame relay switch. Is this likely to ever happen in production environments? Is the configuration done this way solely because this is a lab?

Forsaken_GA

dear god no, not in this day and age. Frame relay is a dying technology, it simply cannot handle the performance today's networks require.

It may have been the case once upon a time if you worked in a service provider environment, but these days, it's pretty much relegated to theory. MPLS is it's replacement, in most cases.

CodeBlox

My professor said the same thing, that it was a dying technology. My question now is: Why isn't it's replacement on CCNA objectives and why is frame relay still a part of the exam objectives? Perhaps because it's still being used in some places?

tomaifauchai

CodeBlox wrote: »

My professor said the same thing, that it was a dying technology. My question now is: Why isn't it's replacement on CCNA objectives and why is frame relay still a part of the exam objectives? Perhaps because it's still being used in some places?

I thought the same not much than 1 month ago and now my answer would be:

Because MPLS look simple at first look but it's heavy under the hood, with all the applications, compared to frame-relay. You can't bring the whole MPLS technology into CCNA, but just the concept around it would be good.
I would say frame-relay is a good first step approach to understand MPLS, though. But that's it. Probably the reason why it's still in the CCNA and it is probably still running in some ISPs or old-hardware equiped companies

Forsaken_GA

CodeBlox wrote: »

My professor said the same thing, that it was a dying technology. My question now is: Why isn't it's replacement on CCNA objectives and why is frame relay still a part of the exam objectives? Perhaps because it's still being used in some places?

Hell, it's still on the CCIE.

Cisco tends to move slow when it comes to pulling out old technologies. They had AppleTalk and IPX on the CCNA long after the protocols had become obsolete. Same with ISDN.

Frame Relay does have some uses, it's one of the few technologies that it's feasible to practice on with subjects like OSPF in an NBMA environment. And frame realy is still in use, but that's primarily because some folks signed up for long contracts (carrier circuit's are not like your cable bill, you don't go month to month), but it is being phased out. I know for a fact that Level3 at least is refusing to renew frame relay circuits once the current contracts expire.

Forsaken_GA

tomaifauchai wrote: »

I thought the same not much than 1 month ago and now my answer would be:

Because MPLS look simple at first look but it's heavy under the hood, with all the applications, compared to frame-relay.
I would say frame-relay is a good first step approach to understand MPLS, though. But that's it. Probably the reason why it's still in the CCNA and it is probably still running in some ISPs or old-hardware companies

MPLS is a huge hack. It's all smoke and mirrors, and the fact that it works at all is what amazes the hell out of me, and why I love it so much.

All in all, I'll be surprised if Frame Relay survives the next time Cisco revises it's exams.

CodeBlox

Oddly enough, I have a Network+ book -which I bought in I think 2008- that at the least discussed MPLS very briefly. I have not much of an idea about the details of it though.

ehnde

Just to expand on this a little, what kind of WAN links are we likely to see at a first "networking" job?

Company A: 400 users and 6 servers in a small town

Company B: 20 users and all managed services in a large city

Company C: 100 users and 5 servers in a large city

Company A datacenter in a large city

alan2308

CodeBlox wrote: »

In the lab, we have to use a router as a frame relay switch. Is this likely to ever happen in production environments? Is the configuration done this way solely because this is a lab?

If you're still using Frame Relay, the FR switches are handled by your service provider. That's why its drawn as a cloud, much like the Internet. You don't really know whats in there, nor do you really care.

So no, you probably wouldn't be dealing with FR switches unless you're working for a service provider.

Forsaken_GA

ehnde wrote: »

Just to expand on this a little, what kind of WAN links are we likely to see at a first "networking" job?

Company A: 400 users and 6 servers in a small town

Company B: 20 users and all managed services in a large city

Company C: 100 users and 5 servers in a large city

Company A datacenter in a large city

Well, what you'll actually get will depend entirely on what folks managed to wrangle out of their boss with a budget, and that can vary widely. There's no set standard, you have to baseline your network to see what kind of bandwidth you need before you actually go buy it.

But in the case of Company A, it'd depend on the type of small town. If we're talking rural hicktown, you'll be lucky to get dialup. Maybe a T1 or T3, if you can convince a phone company to run the circuit out there.

Company B, you'll never find out, unless it's as a consultant. A company outsourcing it's IT and with that few users has no need for a dedicated network guy. Chances are you'll find a business class DSL or Cable connection at best.

C and D are similar, in that it's going to depend on how much bandwidth you actually need, and the infrastructure of the city. In C's case, it may be cost prohibitive to bring anything more than a T3 out to your location, or you may be able to contract for and light up some dark fiber. Maybe you take a MetroE hand off. We have a facility that's about 10 miles away from a major peering point, and we have our MPLS circuits delivered as GigE over single mode fiber. At my old job, we had our facility about 20 miles away from a major peering point, and for the connection from our datacenter back to the peering point, we lit up a dark fiber ring, 80 gigs worth going one direction, 80 gigs worth going in the other. These connections were entirely separate from our internet connections, btw. The sole purpose of these links was to get us back to our core routers housed in the peering points facility. Those routers had other circuits, to a variety of providers for actual transit throughout the world.

The same goes for D if your datacenter isn't located within a peering point. As they say, location, location, location. In D's case, if your datacenter *is* located inside a peering point, then you could potentially see any type of network connection under the sun, it all depends on what carrier you go with, and what service you select, which again goes back to your bandwidth needs and your budget. But typically, you'll find circuits within a large facility being delivered over fiber runs, though in some places, the copper runs are becoming more and more popular.

ehnde

Thanks that was very informative! I would +rep you, but the forums won't let me rep you again for now