My First CCIE

So I got through Lab 10 and scheduled a graded mock lab. In the meantime, I re-read the BGP portion of the book, lab’d out some scenarios and, actually, learned a few things. The day before the graded lab, I was sicker than a dog but what can you do? The show must go on. 6am the next morning, I was starting my lab. Luckily, I felt significantly better.. fever was way down.. aches were almost gone. I can do this. The parts I knew.. I breezed through. There were a few things I had to spend time on. Unfortunately, too much time. I think I set up BGP correctly. The confederation didn’t scare me at all. One path selection task was easy but my solution bit me two tasks later. In a flash of inspiration, I thought I solved it with 1 minute and a half on the clock. Talking to Ethan Banks later, it dawned on me that that solution was bogus too. I knew the answer but it really doesn’t matter now, does it? QoS threw me. I think I got the first one right. The other two were 2 pt questions so they were saved for last anyway.

Surprisingly.. and this was a big deal.. I knew that you didn’t always want to use the same subnet numbers when you lab. It’s too easy to get in a pattern. My network was x.11.x.x but I KEPT typing x.1.x.x. I’d dump a lot of configs from notepad to the devices only to find out it didn’t work. Finding and fixing it wasn’t a problem. But the time it took up was pretty significant and it was a pointless waste. I’ll make MUCH more of an effort to mix it up from now on.

My graded mock lab results are due within 12 hours. I’m not expecting great things. But I do expect that they’ll give me an objective overview of my weak areas. There’s nothing else to do but keep going. I’m going back and hitting some of the previous labs again. My effort to bring my grades up is really an effort to stop making stupid mistakes. I seem to be at the point where it’s not the technology or the command line that’s nailing me the hardest. Much of it is careless mistakes and understanding the requirements. (I’ve also found that sometimes I want to “cheat” the question by implementing “my” solution and working around the task by doing something clever. Wrong.. but clever. Gotta fight that urge.)

So I could keep typing here or go kick some IOS butt. Talk to y’all later.

I’m half way through IE Workbook II. I completed Lab 10 (Difficulty and, while I didn’t do great, I did about average for me. The two areas that gave me the most problems were BGP and multicast. Go figure. But I take some comfort in saying that, other than getting through it so I can take a graded mock lab, my heart really wasn’t in it.

For some scheduling reason, I was thinking that I couldn’t take the graded lab this past Saturday. I set it up for next Saturday. In the meantime, I’m going to go back over some things like BGP.. maybe re-do some labs. When I woke up Saturday morning, I decided to knock out Lab 5 (Difficulty 6) again. I would have liked to pick a harder one but this is the one the roulette wheel landed on so 5 was it.

I haven’t graded it yet but it went amazingly smooth. I even took a stab at the IPv6 BGP steps (which, I understand, are not on the real test at this time). I know there were 5 points I left on the table but I think everything else has a good shot. I have to go back and look at setting up views. I completely chocked on that. Luckily, it was only 2 points. And the other 3 were IPv6 BGP summarization.

There was one 3 point question regarding setting the EIGRP metric weights. I know exactly how to do this but I get messed up on which K value is which metric. I couldn’t find a definitive answer anywhere in the docs that said “K1 is blah, K2 is blah-blah”. Knowing that bandwidth and delay is on by default narrowed my choices but I’m not sure. I’ll find out when I score it today. That’s something I’m just going to HAVE to memorize. During that labs, I’ve taken to only using the Cisco doc pages. If it’s not there, it better come to me in divine revelation cause I won’t have any other resources during the real lab.

The good news is that I had plowed thru almost all of the lab in about 5.5 hours. Another hour to look up some stragglers. And then I took an hour to test and review. I did catch one omission where they told me to advertise some loopbacks and I hadn’t. But… that’s why we check, right?

Wow! It’s been over a month since I blogged! I must be having fun. So.. I just finished Lab 9. This one was interesting in that it didn’t have anything major in it that left me with my mouth hanging open. Sure, I still only made 65.. and I still completely hosed the BGP section but it was mostly mistakes and carelessness instead of outright ignorance. (although there was some). And I think getting a 65 within the 7.5 hour time limit is the highest I’ve made so far. Not passing. But progress.

I have to say that last week was rough. I  just couldn’t get my head out of my.. access-list. I was discouraged and ticked off. I don’t have a clue and there’s no way I’m gonna pass this lab and the problem is that I’m not a quitter so I can’t even do that right. After dumping on Ethan with my anxieties, he explained, “Yup.. you’re in candidate trough. We all pull out of it.” It helped a lot to know that it wasn’t uncommon. I’m pushing thru and focusing on the plan. I have Lab 10 (which I’d like to knock out by Monday) and then I’ll take a swing at a graded mock lab.

I’m coming to terms with the fact that I may not pass the lab. The threat of the open ended questions has just knocked me right off the hope wagon. (which may be part of why I was so discouraged). The lab is a major life challenge as it is. The point of adding the questions seems to be to make it harder… at least for the people who are marginal. (Although a marginal expert.. who’s good enough to pass even if it’s squeaking thru.. still knows a WHOLE lot more than most engineers I’ve known.) Anyway.. the questions are this mystical unknown force. At best, they take 30 minutes off your lab. At worst, they are questions like, “how many milliamps does a 3560 take to push one bit thru a fast-ethernet interface in a 20 millisecond window when there’s 30% utilization on that interface and all other interfaces are in the shutdown state?” or “What is the RFC number that defines the finger service running through IPX encapsulation on AIX?”. Okay.. maybe I’m exaggerating a bit.. but I don’t know that I am. I do know that I can’t memorize (or learn) everything about everything. So, I’m gonna study my labs.. get proficient at what I need to pass the hands-on and hope I get questions that aren’t insane. (Of course, there’s disagreement among candidates and CCIEs alike on what’s insane. The general rule is that if they know the answer, it’s a perfectly reasonable question.)

Either way, I’ll take a shot. And regardless of the outcome, I’m a much better engineer than I was when I started.

Okay.. I just finished going through the Internetwork Expert Class on Demand again. It was worth the time. I picked up a few more things. You know.. when you’re watching it the first time and they’re explaining some little nuance, it’s almost a matter of memorizing the details. After you’ve been through a couple of labs and they talk about those nuances.. well… now it means something and the light comes on. So it was helpful.

In the meantime, I knocked out labs 3 and 4. I did okay on 3. Lab 4, on the other hand… I think I could count what I got right on my fingers. I’m not even going to go into all I did wrong I’m calling it a total loss. I don’t know if I wasn’t focusing or what. But I couldn’t have screwed it up much more if I tried. Nonetheless, I want to keep moving. So, Lab 5 is on deck. With this one, I want to start working on how I lab. I want to focus on the core connectivity and the “low hanging fruit” first and then go back and pick up the little ones. If I’m going to work that way in RTP, I had better start developing that skill now. Also, I want to start watching my points and getting up to 15 points per hour.

The plan…

Work through the labs until I get done with #10. Then take a graded mock lab to gauge where I am. I’ll let you know how Lab 5 works out.

Let’s start off with the good news… last time I took one of the IE assessor tests, I had a problem between two interfaces that appeared to be directly connected. Couldn’t ping from one to the other. And we’re not talking about a hard configuration. It just wasn’t talking. Finally, being stumped, I asked for help in the forums. A couple people responded and brought up other switches. Huh? They’re not connected directly. Oh NO young CCIE to be.. they are connected through a layer 2 tunnel APPEARING to be directly connected. Once you know that’s there, this too is easy. And, once you learn that maybe you should look at the layer 2 diagram every now and then, it becomes easier to see the layer 2 tunnel. Man! A stupid mistake. One of the hard things I’ve had to get over is the resistance to “read the manual” in whatever form that takes. In this case, it’s checking the diagrams. And why should I? I know everything, right? I’m an engineer! But the good news in all of this is that I got it (and got more familiar with L2 tunnels in the process).

The bad news is that I figured out on another IE assessment lab, which dumped me out (terminated me? ended my session?) before I was ready. That’s pronounced, “before I saved my configs and told it to score me”. Five and a half hours I’ll never get back. (six actually. I had spent abut an hour last nite doing prep work… diagrams.. ping scripts, etc. I knew  was taking what could be an 8 hour test in a 6 hour window. Knowing that that hour of prep work would come out of my time if I was at RTP taking the lab, I thought it was a reasonable and fair move). And I was doing… okay. I had gone through everything I knew. Had some problem with multicast but I was mostly there and could come back. I had stepped up to harder (advanced) questions for IP and IOS Features… generally pushing the envelope on each area of study. This is the part I had to go to the Cisco docs for. I was working my way thru them and… goodbye. So.. I have no tangible idea how I did.

But enough of that.. let’s get back to the good news..

As I said, I’ve been going back through the class on demand videos. It’s taking some discipline because I know most of it and tend to drift. (see the first paragraph about knowing it all) But, there are things I’m pulling out of it. Areas that were problems before. I am now the BGP route selection master. Okay.. maybe not master. But I can make it work. I “got it”. What? You want me to make R27 the preferred route to R12’s loopback as advertised into BGP? NO problem! I can do it at the source. I can do it on my boundary. I can make it happen on someone else’s network while drinking a cup of coffee and watching Stargate reruns and instant messaging.

And I’m getting pretty comfortable with MQC. I went back to the book and started at the basics again. This is a CIR. This is your CIR on drugs. This is your CIR sliced up into 50ms segments. Suddenly, it’s getting a little easier. I do need to go back and go back thru “classic” rate limiting. But I’m getting there. Oh.. the other thing that has almost become second nature is OSPF network types over various types of Frame. (For that matter, the frame’s getting easier too.)

I’ve started using a lot of route-maps. And route tags. I’m still not comfortable with redistributing EIGRP and OSPF. It’s not second nature yet so I approach it with some level of fear. So far, it hasn’t been a big problem but I want to really understand it so I can deal with it when it becomes a problem.

So… I’m going to keep cranking on the class on demand. I seem to be getting a lot more out of them the second time around. And.. I’m starting the workbook 2 labs over. I redid Lab 1 the other day and got 1 section wrong. (I failed to put in a virtual tunnel in OSPF. I knew better. I just wasn’t paying attention). Today, I’m hoping to knock out a couple hours of video and maybe start Lab 2 again.

One last thing… I’ve really started using a text editor a lot more. I’m starting to fit it into my style and make it work for me. It was awkward when I first started but I’m finding that workflow. I’ve started keeping a running log of what I’m putting in on what device. I’ll look at a step and write the step and the devices involved. It’s easier for me to keep track of the devices that step is using. For example, if I have to configure ntp on R1, R2, R5, R6, SW4 and SW1, it’s easier to make sure I have all the devices covered if I can look at my text editor and see:

R1

R2

R5

R6

SW4

SW1

If a device doesn’t have a config under it, I didn’t do it yet. It’s easier than having to re-skim the description to make sure I got all of the devices. A snippet might look like…

STEP 1.4
SW3
end
conf t
int range F0/19 – 20
sh
no switch
channel-group 34 mode active
no sh
int Po34
no switch
ip add 129.25.109.9 255.255.255.0
no sh
end
sh etherchannel summ

SW4
end
conf t
int range F0/19 – 20
sh
no switch
channel-group 34 mode active
no sh
int Po34
no switch
ip add 129.25.109.10 255.255.255.0
no sh
end
sh etherchannel summ

.. and I just keep it running thru the lab. I configure what I can in the text editor.. noting the step and the device. Then it’s cut and paste onto the device. The other night when I was working on a lab (in dynamips). I was falling asleep and hosed up a bunch of commands on a switch. I had to reboot the switch to the default config and lost all my vlans (for whatever reason). I thought… huh… if those were in the editor, I could save myself a lot of time and effort. So that’s when I really started using it. Of course, for “one-offs” I don’t usually bother. If I have to set a username on one device, I just do it directly. I think the key is finding out where the balance is for you. But I would recommend trying it.