In the second article in the “Ten Years a CCIE” series, I discuss the Routing and Switching written exam, and the changes to the CCIE exam in the early 2000s.

Passing the written

As was most common in the early 2000’s I attempted my Routing and Switching exam first. Having passed the CCNP exams, my first order of business was to study for the CCIE written exam. Even though I had heard the CCIE written exam was quite easy, I approached it much like I approached a graduate school research project. I didn’t care how easy it was, I wanted to learn the subject material well. I also didn’t want to fail. I had not yet failed a Cisco exam, and I made a pledge with myself: the first Cisco exam I will allow myself to fail will be the CCIE lab.

I assembled a sizable collection of books, including many of the classics. Doug Comer’s Internetworking with TCP/IP, Richard Stevens’ book, Radia Perlman’s Interconnections, and a number of books specifically written for the CCIE exam. I also had many of the books in the Cisco press CCIE collection at that time, including Jeff Doyle’s famous Routing TCP/IP book. As with the CCNP, I practiced many of the theoretical scenarios in the lab just to cement them into my head.

CCIE R/S Written books, 2004

CCIE R/S Written books, 2004

When I took the CCIE written exam, it was harder than I thought. I passed on my first attempt, but only with the score in the 70s. Nevertheless, it was good enough for me to schedule my lab. I passed the CCIE written in January, and scheduled my lab for November, figuring 10 months was enough time to prepare.

Two days to one day

By the time I got around to preparing for the lab exam, Cisco had changed the format rather dramatically. As I mentioned in the previous article a huge part of the mystique of the CCIE exam was that it was a two-day ordeal. In fact, rarely did anybody talk about the test without mentioning the fact that it was two days long. The length seemed an essential characteristic. Thus I, and many of the other CCIE candidates, was quite surprised when Cisco announced that the format was changing to a single day. Although it made the exam seem easier for us, if not in content at least logistically, it seemed that Cisco was changing the very definition of what it means to be a CCIE. Many of those who held CCIE’s at that time were angry about the change, saying that they would not respect anybody who had received the CCIE in the one-day exam.  I myself was quite unhappy with the change, thinking it meant the exam I passed would no longer be well-respected. I wasn’t even sure if it was worth it.

One way that Cisco convinced many of us that the exam would still be worthwhile was to allow some famous CCIE’s to sit for it. I read an article by Bruce Caslow, author of a popular CCIE book, in which he described his experience taking the new format exam.

The author's worn copy of the 2001 edition of Caslow's book

This author’s worn copy of the 2001 edition of Caslow’s book

It was Bruce’s conclusion that the new exam was equally as hard as the old exam. This did something to relieve the distress many of us felt when the exam changed.

Cisco’s stated reason for changing the exam format was that the two-day format was very hard on companies that have to send employees to take the exam. It was entirely possible to fail on the first day. This meant that a company may have paid for two days of travel for an employee who only needed a single day on account of his failure.

I’m sure this reason was partially true. However, I do think Cisco had another motivation as well. The two day exam was successful in keeping the number of CCIE’s fairly low. This is true for couple reasons: first, because of the travel expense, fewer people were able to afford the exam. Second, because the old test was twice as long as the new test, it was simply impossible to administer as many exams. Keeping the number of CCIE’s low was great for CCIE’s, and increased the prestige of the certification. However, it is important to keep in mind that for partners to obtain gold partnership status, they needed a certain number of CCIE’s on staff. The scarcity of CCIE’s made it harder for Cisco to develop their partner network. So, I do think Cisco also changed the test intentionally in order to increase the number of CCIE’s.

Content changes

When they changed the exam to a one-day format Cisco obviously had to drop certain content from the exam. Sensibly, they removed AppleTalk and IPX. These protocols had wide use in the 1990s, but were dwindling by the early 2000’s. They also removed the beginning section of the test where the applicant had to cable his own lab. This was also a good change. A candidate who can configure advanced routing protocol scenarios can probably cable a lab okay, and there were many problems with bent pins. Additionally, the candidate was no longer required to apply IP addresses to his lab; when he entered the room, the basic IP config was already done. He may have to do some additional configuration (to add a GRE tunnel, for example) but the base configuration was there. Again, I agree with this change. If you can configure advanced BGP scenario you can configure IP addresses on interfaces.

More controversial was the removal of the troubleshooting section. Like its two-day format, the troubleshooting section was a defining element of the CCIE lab exam. On the second half of the second day, you went to lunch and the proctors would break your lab. I don’t know whether it was true, but I heard of nefarious mistakes being added to labs. Interestingly enough, in one CCIE prep book I have, the author says that the troubleshooting section was the easiest, claiming that if you spent a day and a half configuring your lab is pretty hard for the proctor to introduce something you won’t notice. Another book I have by a CCIE claims that the troubleshooting section was most difficult. Either way, it was gone.

More changes on the way

An additional concern for me in studying for the lab was that a number of changes were announced which were to take effect in January 2005. Since I was taking the lab in November 2004, it seemed unlikely that if I failed I would be able to take the test again before the changes took effect. This gave me a sense of urgency; I felt that I had to pass. If I didn’t I would be facing a very different exam than the one I had studied for.

The planned 2005 changes were as follows: first, certain technologies were being removed. These included DLSw+, a protocol used for transporting non-routable protocols across an IP network, mainly used for NetBios and SNA. Removing this made sense as it was a small part of the exam and not a heavily used protocol at that time. They also removed ATM. ATM was widely deployed at the time they removed it, but I still think this move made sense. The Routing and Switching exam is focused on enterprises, and enterprises generally did not use ATM as heavily as service providers. In addition, on the Routing and Switching exam, you only had to configure one end of an ATM link. Thus, it was not a major part of the exam. They also decided to remove Voice over IP. Another good idea. Although Voice over IP was in its ascendancy in 2004, Cisco had its own separate exam for VoIP, and configuring a couple of dial peers seemed out of place in an enterprise routing and switching exam. ISDN was also gone, and again, this was correct. ISDN was waning in 2004. There was no sense studying these dying protocols.

There was also a major subject added to the exam: IPv6. This made me very nervous. When I took the exam in 2004, IPv6 was nowhere on it. It seemed to me a huge subject, and I was very afraid to fail if I had to retake the exam in 2005. I have many complaints about Cisco’s administration of the CCIE exam, which will follow in my reflections. However, I think they generally do a good job of updating the test. All of the changes they made in 2005 were perfectly reasonable given the state of technology at that time.

Exam changes are a perennial problem for candidates.  We all dread the introduction of new topics which can broaden the amount of material we have to master.  For people dependent on test preparation outfits like INE, the addition of new material can be especially nerve wracking because it takes time for those guys to update their material.  This pressure often makes candidates try extra hard to pass the exam before the announced changes take effect, which leads to another problem:  now you are studying old technologies.  In my case, I didn’t want to get into new subject matter, but on the other hand, did it really benefit me to study DLSw and ATM?  Regardless, as a candidate you need to be ready for announcements that will change the course of your study, and you need to plan out what will happen if you fail the lab shortly before the changes go into effect.

In the next article in the series, “In those days, you had to build a lab,” I discuss building a CCIE lab in the early 2000’s, including why I felt it was important to have a home lab, and what I had to do to assemble it.