Starting CCIE Written. Just had "the talk" with Mrs. Zartan.

Zartanasaurus

The most important thing was that she approved. I told her the amount of time and work that would be involved in getting all the way to the CCIE lab and she supports me 100%.

Now with that hurdle out of the way, here's what I've been up to lately.

I read Interconnections by Perlman. Great book BTW, despite its age. Her writing style is just perfect. I don't think I've ever read a better explanation of spanning tree or link-state protocols.

Currently reading TCP/IP Protocol Suite by Forouzan. I know Internetworking by Comer and TCP/IP Illustrated by Stevens are the standard, but I chose the Forouzan book for a couple of reasons. 1) It had an ebook edition. 2) I was doing some research on books used by universities that offer Master's Degrees in Networking and I found this book being used at DePaul.

I bought the 2nd edition of TCP/IP Illustrated for Kindle and I just thought it was terrible. Nothing like the original which was well written and easy to understand. I gave it a good 5 or 6 chapters before I called it quits. I think a reviewer on Amazon hit it on the head when he said "I disagree with the other reviewers who state that Fall retains the excellent writing style of the original. Whereas Stevens is known for succinct, clear prose that covers topics in a straightforward, readable way, Fall seems to have felt that adding verbosity was a necessary step in adding additional topic coverage". I read about 1/3 of the 1st edition then stopped in anticipation of reading the new updated version. I will probably go back and finish at some point.

The Forouzan book is well written with clear diagrams that aid in your understanding. I don't have much to compare it to with other books of its kind, but it's better than the TCP/IP Illustrated 2E. Has excellent reviews on Amazon as well.

Other book I'm currently reading is Routing TCP/IP Vol I. Only on Chapter 5, but I love it so far. I can see why it's so highly recommended.

I also have the following books:
Routing TCP/IP Vol II
Internet Routing Architectures
Developing IP Multicast Networks
MPLS & VPN Architectures
MPLS Fundamentals
BGP Design & Implementation

I've seen that the Cisco QoS Exam Guide and End to End QoS Network Design books come highly recommended, so I will pick those up eventually.

I also plan on getting an all access pass from INE. I've read that the ATC plus good reading material will get you through the written.

Any other books I should be picking up for either the written or lab? I've seen Cisco LAN Switching recommended a lot. Preference or Kindle or other ebooks over hardback only, but if it's a must have it's a must have. I see Cisco has done a good job converting lots of their older material into Kindle format if it comes highly recommended for CCIE studies.

My study plan is to go through Routing TCP/IP, a QoS book and an MPLS book then watch ATC/R&S Bootcamp videos on the topics I read about. I also plan on building the INE CCIE lab with a combination of GNS3 and 3560s for labbing.

Anything I'm missing?

jamesp1983

Good luck man!

Forsaken_GA

You've hit everything I would recommend. The only other advice I'd give is this -

Don't fear the Written. It is a pathetically easy exam that any CCNP with a year or two of experience can pass. I cannot express my disappointment in the Written's difficulty.

The only thing to keep in mind is that passing the Written puts you on a clock. You have 18 months from the date of passing to go ahead and take your first attempt, otherwise, you wasted $300 bucks. If you don't believe you'll be ready to attempt the lab within 18 months, then hold off on the Written until you do... but once you do, pull the trigger. Nothing motivates you like being on the clock.

aldous

troubleshooting ip routing protocols

A real good book, slightly old (it has an igrp chapter) but really helped in my understanding of various protocols

Zartanasaurus

Off to a quick start. Using Petr's Study tips to plan my path. Have been documenting what I do each day in Google Calendar and scheduling repetitions based on Supermemo.
Also bought 3x HP NC340T 4-port NICs for my server. Hope to have everything by Thursday at the latest, then it's super lab time!

Saturday:
About 10 hours total, although a lot of that is time spent goofing off on the internet. Had to do laundry, eat dinner and do dishes too of course. I really need to get that under control so I can have more focused sessions.

Read Chapter 5 & 6 in Routing TCP/IP, Vol 1 (RIP v 1 & 2)
Completed all Review Qs, Config Labs and Troubleshooting Qs in Chapter 5.
Completed all Review Qs & Troubleshooting Qs in Chapter 6.
Scheduled Repetition 2 for 4/11.

Sunday:
A little better with being focused. Spent about 6 hours total with about 3 hours of real studying time in-between breaks and goofing off.
Completed RIPv2 Config Labs.
Scheduled Rep 2 for 4/12.
Read Chapter 7 (EIGRP)
Definitely a big chapter. Chapters 5 & 6 on were 100 pages combined. Chapter 7 was almost 100 by itself.
Highlight of the night was figuring out how to read that DUAL finite state machine chart. :lightbulb:
Didn't feel like starting on the labs, and questions so I'll do that tomorrow. Looks like OSPF is ~150 pages.

I've been saving my baseline lab configs from the Doyle Book in GNS3 so I can repeat them on schedule. It would be a great feature for books in the future to contain downloadable GNS3 configs to do troubleshooting and config exercises on. Takes a bit of time to get everything up and running just so you can start doing the actual lab. I tried googling around to see if anyone had already made topologies out of the end of chapter labs, but didn't come up with anything. I can make them available if anyone wants them in the future.

Once I get to repetition 3 on a topic, I will include new questions and labs. According to Kindle, I'm 36% of the way through this book.

bermovick

Zartanasaurus wrote: »

Saturday:
About 10 hours total, although a lot of that is time spent goofing off on the internet.

Sunday:
about 6 hours total with about 3 hours of real studying time in-between breaks and goofing off.

This is why I eventually stopped keeping track - whether I get sidetracked to go cook, or hop onto a game for 30 minutes or do something at work when I study at night, I never was sure how much I really studied. Better not keeping track than have wildly inaccurate numbers, I felt. Not that I'm suggesting anything by it; it just didn't work for me. Now I feel less pressure I guess to have X hours or whatever

NOC-Ninja

You can fix that **** through slimtimer.com
I pause it when I start browsing or get out of mychair and start doing something else.

Zartanasaurus

NOC-Ninja wrote: »

You can fix that **** through slimtimer.com
I pause it when I start browsing or get out of mychair and start doing something else.

Awesome, thanks!

QHalo

Not waiting for the DC stuff? Good luck either way! Keep the thread updated.

Zartanasaurus

QHalo wrote: »

Not waiting for the DC stuff? Good luck either way! Keep the thread updated.

The test for this isn't out until September, and some of the books on the reading list are being updated to 2nd editions. Then it will probably be another month before the major training vendors get their guys to pass this test so they know how to teach it.
The DC stuff has lots of applicability to my day-to-day duties, so I'll still stay current and try to sneak some reading in here and there when I can. The discussion of load balancers came up in a meeting today, so if we purchase an ACE, that's one more thing on the lab equipment list I could play with as practice.

Zartanasaurus

Did all the end of chapter questions and config labs for EIGRP. Very disappointed in the config labs. They were straightforward and very vanilla. The RIP labs were harder! The ROUTE labs Jeremy did on CBT Nuggets were 50x more involved. They have two different CCIE lab packages, and I have an all access pass through work, so I may start supplementing with those.

Instead of going with a software timer, I broke out the good old fashioned wrist stopwatch. For some reason, the fact that I knew I was timing myself kept me from goofing off at all. I got 4 hours of reading in with a couple 5 min breaks. Amazing that I only managed to read 60 pages on OSPF in 4 hours, but I take lots of handwritten notes. I feel that learning retention is better that way vs typing notes. The good news is, the remaining 100 pages are mostly graphics of the LSA packets and some basic how-to configuration stuff. I just got done reading the Interconnections book and I took tons of notes on LSA packet types and OSPF in general. I think the next 100 pages will go much faster.

Zartanasaurus

Day 4:
Just under 6 hours today. Timer method is keeping me focused and distraction free.

Completed repetition 2 for Dynamic Routing Protocols Basic, Chapter 4 Review Questions. Scheduled repetition 3 for 4/17
Finished Chapter 8 on OSPFv2 and did review questions. Scheduled repetition 2 for 4/14

Lots of material to absorb in this chapter. Glad I split the reading up over 2 days. I'm definitely going to be going back over my notes a lot to stay fresh.
I really like Doyle's method of going into a deep dive on the technical details and showing you how it all works together without doing configurations.
When you're done you can sit there and write out on paper what the routing tables, neighbor tables, etc should look like before you even look at a show screen. Then you get some configuration info and case studies showing you how to do what and when.

I started changing the colors of the event based on repetition #. Red = 1, yellow = 2, green = 3. Looks much nicer on the calendar and you can see your progress better. My first repetition #4 isn't until the 29th, so I have plenty of time to pick a 4th color.

Tomorrow is OSPFv3 which looks to only be about 40-50 pages. I wasn't keeping the most accurate time on my reading before Monday, but I'm going to give myself 10 hours for those days, plus the 10 hours I've done today and yesterday. So 20 hours total. This is like taking on a part-time job.

Roguetadhg

Good luck! I'll keep tabs on this one

Zartanasaurus

Day 5:
One of my quad NICs arrived. Other two should be here tomorrow.

4.5 hours.

Repetition 2 for RIPv1 Questions and labs.
Repetition 2 for RIPv2 Questions.
Repetition 3 scheduled for 4/18

Repetition 1 for OSPFv1 labs.
It's a big lab but not particularly difficult. He doesn't come right out and tell you exactly what to configure in every case. Sometimes you just have to look at it and know what's required. It takes a long time to get setup the first time in GNS3 just because of the amount of routers and he doesn't give you a diagram. All you get is a list of routers, their interface IPs and what areas they belong in. It took me longer than I would've liked because of some stupid mistakes, IE trying to configure virtual-links on the wrong set of routers.

I didn't do any reading other than reviewing my RIP notes after I finished the lab. Tomorrow is Rep 2 for the RIPv2 lab and the chapter on OSPFv3. After that, there's only 150 pages of content left, excluding IS-IS. I should probably be done sometime Friday.

Zartanasaurus

Day 6 - 2 hours
Didn't post my progress last night, but I did make progress. The last couple days have been tough to get any reading done at work.
Repetition 2 for the RIPv2 Lab. Spent a lot of time trying out a lot of different ways to do the lab other than the solution along with some general piddling around. When I got home I ended up going right back out again for a welcome distraction for a hobby I haven't kept up with much for the last year or so. Came home late and read a little bit of the OSPFv3 chapter then just called it quits.

Day 7 - 3 hours
My last two quad NICs are here. Put them in and my server wouldn't boot. And neither would one of my 3560s. Should've plugged it in first before I hauled it out of storage I guess. Gonna mess with it again on Monday since it's not a requirement that I have the INE topology ready just yet.

Repetition 2 for the EIGRP review questions and lab. Like the RIP lab I spent some time toying around and just getting more comfortable with the show commands.

Not a lot of meat in the OSPFv3 chapter. I didn't take as many notes as I had been taking and I didn't even bother with the lab as it was very simple and straightforward.

Tomorrow I'm going to break out my whiteboard and just do a comprehensive review on everything I've learned up to this point to reinforce some things. Then it's the last 4 chapters in the book, which should go by rather quickly. My next book is going to be Cisco LAN switching. Want to reinforce what I read in the Interconnections book to keep it fresh in my mind.

Total - 29.5 hours

TheShadow

NOC-Ninja wrote: »

You can fix that **** through slimtimer.com
I pause it when I start browsing or get out of mychair and start doing something else.

A truly useful site. Thank you for that +1

@Zartanasaurus good luck on your journey

Zartanasaurus

Day 8 - 7.5 hours
Total - 37 hours

As planned, started off with a comprehensive whiteboard review. First two hours covered RIPv1/v2 and EIGRP. I just dumped everything I could think of about a particular subject on the whiteboard. Then I'd skim the relevant sections and write out what I missed. If I thought it was particularly important, I added it to my review questions so I'll see it in the next repetition.

Took a short break, then did repetition 2 for the OSPFv2 Review Questions. Then I did another comprehensive whiteboard review on OSPF which took up the next two hours. Spent a lot of time checking multiple sources to understand the no-redistribution, default-information-originate and no-summary commands and how they affect NSSAs in various combinations.

Last 3.5 hours spent on Chapters 11 & 12 for redistribution and default routing/ODR topics and review questions. Nothing too deep in these chapters. Repetition 2 scheduled for 4/18.

I feel like the whiteboard session helped a lot and allowed me to catch things I had missed the first time around. I think I've decided on the order I will go from here on topics after I finish Volume 1 of Routing TCP/IP.

Cisco LAN Switching
Routing TCP/IP, Volume 2
Developing IP Multicast Networks
MPLS Fundamentals or MPLS & VPN Architectures
Cisco QoS Exam Guide
Troubleshooting IP Routing Protocols

INE doesn't have a suggested book for MPLS, so I'll have to look into which one of those might be preferred over the other.

Forsaken_GA

Zartanasaurus wrote: »

INE doesn't have a suggested book for MPLS, so I'll have to look into which one of those might be preferred over the other.

If you bought the All Access Pass, then just go for their MPLS course. I tried to start with MPLS Fundamentals and MPLS and VPN Architectures, but they just didn't work too well for me. The INE video course, otoh, worked beautifully and brought it all together, even if I did want to smack the crap out of Keith Barker every now and again.

Zartanasaurus

Day 9 - 6.5 hours
Total - 43.5 hours

The stopwatch method continues to work wonders for me. I wish I had started doing this before.

Completed repetition 2 of the OSPFv2 config exercises and scheduled rep 3 for 4/22. This lab inspired me to find out how to do loop of pings in TCL. Easy find on the google and it saves a lot of time when you're trying to verify basic connectivity to 30 different IPs.

Concluded the TCP/IP Routing book with Route Filtering and Route Maps. Short chapters of mostly case studies. One of the config exercises is screwed up in the Kindle version where it doesn't show you the interface IPs, then you're supposed to configure route filtering. Kind of hard to do when you don't know what you're supposed to be filtering.

Bought the Cisco LAN Switching Kindle book and read the intro. Hopefully it doesn't have the same formatting issues. Can't wait to get started all over again tomorrow.

Zartanasaurus

Got another 3560 to replace the one that wouldn't boot so I'm good as far as that goes. I reseated the NIC cards in the server which fixed the no booting issue, but I got an error message about an unsupported PCI card. Turns out, this is a known issue with HP servers running a specific AMD chipset.
HP ProLiant DL385 Server series -  Advisory: NC340T PCI-X 4-Port 1000T Gigabit Server Adapter must be Installed in Slot 1 of HP ProLiant DL385 when Two Emulex LightPulse LP10000 HBAs are also Installed in the Server - c00709385 - HP Business Support

Now I have to find a 100mhz PCI-X quad port card to replace one of them with. It's always something.

Zartanasaurus

Day 10 - 7.5 hours
Total - 51 hours

Got a great day of reading in at work. If I'm lucky I can get about 3 uninterrupted hours in starting from an hour before lunch. Started the Cisco LAN Switching book and went through the first 7 chapters. The beginning isn't much and I mostly skimmed through it. Chapters 5 & 6 contain a little more meat. I still like Perlman's explanation of spanning-tree better, but Clark goes into the Cisco-specific options of backbonefast, uplinkfast, portfast. Other than exception configuration BPDU processing, there wasn't anything there I hadn't read elsewhere. The end of chapter config labs are easily answered in your head, so no need for any real labbing just yet.

Lots of stuff missing here. May have been better to just go back and read the CCNP Switch FLG again.

Looking ahead, it only seems like chapters 8, 11-13,16 need to be read. Has anyone read any of the design and implementation chapters?

Zartanasaurus

Day 11 - 6 hours
Total 57 hours

I fell asleep while reading last night. One minute I was reading about MLS, the next thing I remember is thinking "oh crap I fell asleep". I don't even remember feeling tired or putting my head down. I guess that's what happens when you read in bed sometimes.
I covered chapter 8 on Trunking then got most of the way through Chapter 11 on Layer 3 Switching. I tried to finish after I awoke, but it just didn't feel the same so I called it quits. The end of chapter questions/config exercises don't seem to be adding much value. It's still a pretty good book where the coverage is relevant. I should be able to finish the last chapters on VTP, multicast and Troubleshooting today. I will skim through the design chapters at first and revisit if they look good.

Zartanasaurus

Day 12 - 5 hours
Total - 62 hours

Since I fell asleep last night I wasn't able to do my 2nd repetitions for default routing and redistribution. Completed them first thing this morning at work, as well as rep 3 for the RIP v1 lab.

Finished the last few pages on MLS then into Chapter 12: VTP. For the last few chapters, I've been reading the 3560 config guide and taking notes, then going back to the book for review. I think this is helping a lot. I need to backtrack and read the sections on spanning-tree tomorrow then start on Chapter 13: Multicast services.

Zartanasaurus

Forsaken_GA wrote: »

If you bought the All Access Pass, then just go for their MPLS course. I tried to start with MPLS Fundamentals and MPLS and VPN Architectures, but they just didn't work too well for me. The INE video course, otoh, worked beautifully and brought it all together, even if I did want to smack the crap out of Keith Barker every now and again.

If I'm not mistaken you have a background with MPLS, right? Would you still recommend the same for someone w/ no MPLS experience?

Forsaken_GA

Zartanasaurus wrote: »

If I'm not mistaken you have a background with MPLS, right? Would you still recommend the same for someone w/ no MPLS experience?

When I was studying for the MPLS exam, I did not have a background with MPLS (I do now), I had a strong BGP background already by that point, and that certainly helped, but it wasn't until I sat down and watched the video series and saw it all put together in a consistent and logical manner that MPLS started to make sense.

Zartanasaurus

What am I missing here?

RTA, B, C, D are running RIPv2
RTC, D, E, F are running OSPF, area 0
So RTC, D are ASBRs.

Networks 192.168.1.0/24 and 192.168.2.0/24 are loopback interfaces on RTA, B representing a LAN segment. C & D are both doing mutual redistribution. No kind of filtering.

RTC is seeing the loopbacks as O E2 routes, like I expect. RTD is seeing them as R routes, which I don't expect.

When I do a sh ip ospf da, C sees D as the ADV router. D sees itself.

Forsaken_GA

Well, before I get into any explanation... what *were* you expecting?

Zartanasaurus

Forsaken_GA wrote: »

Well, before I get into any explanation... what *were* you expecting?

I was expecting to fool both C & D into thinking the optimum path to the RIP networks were through the OSPF domain, creating a loop. Looks like I got halfway there.

Forsaken_GA

Zartanasaurus wrote: »

I was expecting to fool both C & D into thinking the optimum path to the RIP networks were through the OSPF domain, creating a loop. Looks like I got halfway there.

Ok, here's why that's not going to work -

Remember, redistribution is done from the routing table, not the associated databases.

So both routers are learning the path to the loopbacks via RIP, since they're both in that routing domain. If they both redistribute RIP into OSPF, logically, it would follow that they install the OSPF routes due to administrative distance.

However, if they did that, they would no longer have a RIP route to that prefix in the routing table... which means they couldn't redistribute it into OSPF, so there'd be no OSPF route to overtake the RIP route. It's a chicken and egg problem. In order for either router to have the routes available in the instance of mutual redistribution that you have above, one of them *must* have the RIP routes.

The answer lies in how OSPF treats external information. Remember, OSPF is link-state within a given area only. Routes that are between areas or externally behave as though it were distance vector. And this means split-horizon comes into play.

Simply put, OSPF will not advertise a route that is external to it's area back into the area it learned it from. So when you redistributed RIP into OSPF on RTD, RTC learned the routes via it's area 0 link. It will not then advertise the same routes back onto it's area 0 link. I'll bet if you removed redistribution on RTD and then reapplied it, you'd see the RIP routes on RTC and the OSPF routes on RTD.

Now, I'm not sure whether or not you need this, but for anyone else following along - accepting those routes via OSPF is the incorrect thing to do in the first place. You're getting the routes via a natively connected routing protocol, and native routes should *always* be preferred over externally injected routes if you want to avoid sub-optimal routing. There's a few ways to avoid this, you can play with AD, or use route-tagging. In this situation, I would personally raise the AD of External OSPF prefixes to 121 on RTC and RTD in order to guarantee that the RIP routes took precedence over the external OSPF routes (thereby matching the behavior of EIGRP).

Zartanasaurus

Forsaken_GA wrote: »

Ok, here's why that's not going to work -

Remember, redistribution is done from the routing table, not the associated databases.

So both routers are learning the path to the loopbacks via RIP, since they're both in that routing domain. If they both redistribute RIP into OSPF, logically, it would follow that they install the OSPF routes due to administrative distance.

However, if they did that, they would no longer have a RIP route to that prefix in the routing table... which means they couldn't redistribute it into OSPF, so there'd be no OSPF route to overtake the RIP route. It's a chicken and egg problem. In order for either router to have the routes available in the instance of mutual redistribution that you have above, one of them *must* have the RIP routes.

The answer lies in how OSPF treats external information. Remember, OSPF is link-state within a given area only. Routes that are between areas or externally behave as though it were distance vector. And this means split-horizon comes into play.

Simply put, OSPF will not advertise a route that is external to it's area back into the area it learned it from. So when you redistributed RIP into OSPF on RTD, RTC learned the routes via it's area 0 link. It will not then advertise the same routes back onto it's area 0 link. I'll bet if you removed redistribution on RTD and then reapplied it, you'd see the RIP routes on RTC and the OSPF routes on RTD.

Thanks. Fantastic explanation as always. I'd pay money to work with you for a year.

Now, I'm not sure whether or not you need this, but for anyone else following along - accepting those routes via OSPF is the incorrect thing to do in the first place. You're getting the routes via a natively connected routing protocol, and native routes should *always* be preferred over externally injected routes if you want to avoid sub-optimal routing. There's a few ways to avoid this, you can play with AD, or use route-tagging. In this situation, I would personally raise the AD of External OSPF prefixes to 121 on RTC and RTD in order to guarantee that the RIP routes took precedence over the external OSPF routes (thereby matching the behavior of EIGRP).

This is exactly what I was doing. I was trying to build a lab with a loop so I could fix it with route-tagging or the distance command.

Zartanasaurus

Day 13 - 6 hours
Total - 68 hours

Repetition 3 for RIPv2 lab and questions. Time to add some additional info for future repetitions. Finally understand the small bits I was missing here and there.

Repetition 2 for Route Filtering and Route Maps. As you can see, I spent a little time trying to figure out why I the lab wouldn't break the way I was expecting. I started in on the Multicast & IGMP section of the Clark book. Finished up through CGMP which I had no knowledge of before. I had enough multicast/IGMP snooping knowledge to enable it here for deploying OS images over the network, but I didn't have a super deep understanding. A good portion of what I read today was review, but the section on CGMP was very interesting. It's a very clever way of building the database without the switches having to support IGMP snooping.

Tomorrow is EIGRP rep 2 and spanning tree rep 1. I'm going to read over the all the sections in the 3560 guide pertaining to spanning-tree so I don't expect to get much else done reading wise tomorrow other than finishing the multicast chapter. If I finish that, I'll supplement with the Cisco docs again. Really liking that particular learning method.

It's unlikely that my replacement quad NICs will be here tomorrow, so it will be Monday night at the earliest that I can run the INE topology and start on their material. I should at least start with the Written boot camp videos just to add another source of info into my repetitions.

Boss is also trying to arrange for me to go to some Nexus courses through Firefly or Global Knowledge, so I have that on the horizon too.