Frame (Relay) of Reference


A while back I wrote a CCIE lab-related post about RIP and why it’s still on the lab.  Most of the questions that I see lately revolve around another old technology and the curiosity of why it’s still contained in the vaunted lab blueprint.  I speak of frame relay.  The much-maligned WAN technology that no one seems to be able to explain the reason for existing anymore, yet the weary lab candidates find themselves pour over at the last minute.

Frame relay has been around forever.  And slightly shorter than forever, it’s been on the CCIE lab blueprint.  I’ve been actively studying for my exam since July 2008.  And as far back as I can remember, I’ve been studying frame relay.  And ever since that time people have been asking why it’s even on the exam.  I’ve heard people say it’s too old.  It’s past it’s prime.  No one offers it anymore.  Recently, with the change to the version 4 lab exam, MPLS was introduced as a configuration topic.  We all thought “Surely frame relay will be gone with its apparent successor on the lab now…”  And yet, frame relay is still there.  In fact, they added frame relay switching back onto the blueprint.  That was a huge change for me, never having dealt with that side of things before.  So, here is frame relay configuration of all kinds taunting us with is outdatedness.  Forcing us to keep up with our grandfather’s WAN technology.  And, in my opinion, that’s just the way Cisco likes it.

Just like RIP, frame relay wasn’t necessarily meant to be on the exam as a test of your ability to map IPs to DLCIs.  Frame relay introduces a whole level of configuration possibilities to the lab to test your wits without being obvious.  The whole point of keeping it on the lab isn’t to make you expend effort in getting 2 or 3 points.  It’s laying a foundation that could cost you 10 points or more.

I remember reading the Building Scalable Cisco Internetworks (BSCI) routing exam guide a few years back studying for my CCNP.  It was my first real exposure to the depths that routing protocols could go to.  In one the chapters over OSPF, there was a section that covered a ton of information on running OSPF over non-broadcast multi-access networks.  There were tables and charts and graphs galore.  All dedicated to one sub-topic.  Why was that?  Because the intricacies of running OSPF over frame relay are legion.  There are multiple network types that determine DR and BDR functionality and hello timers.  It’s like a mix-and-match sale at the department store.  So many fun combinations from so few parts.  That chart was just the tip of the iceberg, though.  Because once you’ve gotten a frame “cloud” in your lab, the real fun can start.

Suppose you have a really simple question in the lab like this:

Configure frame relay between R3 and R6.  Do not use static
mapping or inverse ARP. (2 points)

Straightforward, eh?  You could probably nail up this frame configuration in about 5 minutes.  Well, think about the rest of this example lab and find all the things that rely on Frame Relay to award points.  Let me give you a few examples:

Configure OSPF on all routers indicated in the diagram.  
Between R3 and R6, use a network type the provides for the
fastest recovery times without a DR/BDR election. (3 points)

Configure the link between R3 and R6 so that routing protocol
packets are never dropped.  If the queue has more than 20 packets,
reduce the bandwith to 32kbps. (3 points)

Configure the link between R3 and R6 to authenticate using CHAP.
(2 points)

Now, I’m not saying that all of those questions could end up in your lab.  But think about it like this:  That single 2-point FR question just became a 10-point chunk of your lab.  Now, if you don’t configure your FR section correctly, you are halfway to failing the lab from one misconfiguration!  Remember, no partial credit means that even if your whole OSPF section is right, failing to get FR working correctly means the script used to grade your exam won’t see all the routes on R3 and R6, and so you won’t get any points for OSPF.  It also means that your QoS won’t work, so there goes three more points.  And so on.  Soon, you could fail the lab simply because that outdated technology you didn’t really care about sucker punched you.

I’ve always held the belief that the CCIE lab is designed in a way to test every facet of being an network superhero.  Not just the obvious things like OSPF configuration or BGP troubleshooting, though.  As I’ve said before, the lab manual is carefully written by a team of highly competent people.  There are no extraneous words, and there are no unimportant tasks.  It’s like a game of Jenga.  The things you do at 10:00 a.m. have a direct impact on the last task you do at 4:00 p.m.  If your lab isn’t built carefully and solidly like a Jenga tower, it will topple down on the desk in front of you and knock your cup of colored pencils and markers onto the floor (metaphorically, of course).  Like it or not, frame relay is one of those blocks that forms the foundation of your tower.  Not because Cisco is protecting their investment in legacy technology.  Not because they are trying to wear lab candidates out with archaic configuration tasks.  Frame relay still exists on the blueprint because it teaches candidates to get their configurations right in that critical first hour of the lab because the Jenga tower that is your lab exam depends on each and every part.  And since frame relay can touch so many parts of the test, Cisco likes to keep it around to make sure you’re paying attention.

So the next time you find yourself staring at the blueprint and asking yourself “Why does Cisco still put frame relay on the lab?” you might try asking instead, “What does Cisco want me to learn from configuring frame relay?”  I think you’ll find the answer to the latter question a lot more enlightening.  With all of the things that depend on a simple frame relay link, one miscue could be the difference between a detailed score report and a simple 4-letter one.  Because from Cisco frame of reference, frame relay isn’t just a simple test question.  It could be one of the most important topics on your exam.

6 thoughts on “Frame (Relay) of Reference

  1. Pingback: Tweets that mention Frame (Relay) of Reference | The Networking Nerd -- Topsy.com

  2. I just finished up with Frame Relay labs in my prep for the R/S lab later on this year. I guess I never really thought about it in the way you just explained in your post. Makes perfect sense though.

    I was moaning about FR at lunch a few weeks ago with a CCIE I know and another network guy and they both started enumerating a bunch of different things that FR is still used for. Like CatOS, FR is going to take a long time to go away.

  3. Your explanation certainly makes sense, but I don’t agree that it validates the inclusion of frame relay on the exam. As you explain, frame relay configuration affects a disproportionately large chunk of your overall score. But why? What’s the point? If I know every inch of OSPF, why should I penalized in that area for being rusty in another?

    I have never seen frame relay implemented on a modern network. I never hear people ask about it outside the context of a certification. It seems most people who do come into contact with it on a production network are there to replace it.

    In my opinion, the obscure frame relay topologies are akin to nailing the keyboard to the ceiling: it makes the exam more difficult, but doesn’t test any relevant knowledge. Cisco would do better to test much more realistic scenarios, and we all know there is no shortage of challenging situations one comes across in the real world even without clinging to legacy technology.

  4. Stretch,
    I’ve always held that the frame relay portion is perfect for Cisco to test how well you react to configurations being influenced by other factors in your lab. It’s much like what used to happen when the proctor ‘broke’ your lab at the end of the first day of the 2-day CCIE lab. They are using the question in an effort to force you to improvise when stuff starts breaking. Especially if you forget to put a “broadcast” keyword on your mapping or you leave PPP encapsulation out of the equation.

    Now, I agree wholeheartedly that frame relay as a technology is probably way past it’s prime. I’ve never encountered it in reality, and the one time I was asked about it I laughed and told the customer to go ask for MPLS instead of “a WAN technology better suited to Michael Jackson or Miami Vice.” But, like we are constantly reminded, the CCIE isn’t about how we’d do things in the real world. It is the imperfect scenario-based solution for a team of test writers to gauge our knowledge about minutia of configuration. And until they find a way to make MPLS just as insidious as frame relay, we’re probably going to be stuck with it on the v5 lab as well. But by that point, we’ll have to setup our lab in a museum to find technology capable of running frame relay.

  5. Rumors say Cisco sticks to frame-relay simply to catch candidates off-guard on various point/multipoint scenarios 🙂 Just look at OSPF.

    FR encapsulation is still being used for circuits that connect customer egde routers to MPLS. Also Ethernet, ATM, DSL, SDH, BRI/PRI … what’s available from local telco to connect router to nearest POP. But that’s point-to-point that ends up on PER interface, last I saw let’s say multipoint on true FR network was … heh many years ago. Though it would not surprise me if archeological research showed some remainings of it still working somewhere. Together with SNI and Novell printer services (yuck).

  6. Pingback: CCIE Version 5: Out With The Old | The Networking Nerd

Leave a Reply to stretch Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s