1st CCIE lab attempt blog and help for candidates.

Turgon

Hi,

My previous thread entitled 1st CCIE lab attempt starting April 2007 was temporarily offline so this is a new thread covering my efforts.

I'm working evenings and weekends to prepare for my belated first CCIE lab attempt in Spring 2009 (bumped from July 2008 due to work/family commitments). This thread will be a record of my progress. I hope you will all drop by with encouragement. I will certainly gladly offer advice to anyone else working towards the lab. I also work fulltime.

I have been working in the biz for 10 years and have finally gotten around to finishing my lab prep this year after a few years concentrating on my work in the field. So Im returning to the CCIE now and pulling on my previous years of experience with cisco, cisco studies and professional networking.

I will post up my findings on the latest lab I have just completed at home shortly.

Meanwhile, drop by and say hi

dtlokee

Is this your first attempt? I have 4 days to decide if I am going to pay for my lab or not, I just never feel prepared enough to go for it, but I need to jump back on the horse at some point.

sprkymrk

Hi Turgon! I'll be interested to see your methods and how you progress. Thanks for starting a new thread and I wish you great success in passing on your first attempt.

Although I am not unfamiliar with Cisco, it's certainly not my forte. It's a skill that I hope to develop more in the near future. I am sure I'll learn a lot by lurking on this topic. Thanks!

ajs1976

Hello Turgon, good to see you around here.

Turgon

sprkymrk wrote:

Hi Turgon! I'll be interested to see your methods and how you progress. Thanks for starting a new thread and I wish you great success in passing on your first attempt.

Although I am not unfamiliar with Cisco, it's certainly not my forte. It's a skill that I hope to develop more in the near future. I am sure I'll learn a lot by lurking on this topic. Thanks!

Well thank you and drop by anytime. I will certainly try and help out!

Turgon

ajs1976 wrote:

Hello Turgon, good to see you around here.

Hello Andy,

Great to hear from you again my friend. If you are in touch with any of the guys from icertify etc be sure to ask them to drop by!

Turgon

dtlokee wrote:

Is this your first attempt? I have 4 days to decide if I am going to pay for my lab or not, I just never feel prepared enough to go for it, but I need to jump back on the horse at some point.

Hi dtlokke,

Yes this will be my first attempt, however I'm no stranger to cisco or the CCIE process. I passed the CCNA in October 1999 and took the CCIE written in 2001. I have followed groupstudy for many years since then and did a lot of prepartion for the lab in 2002. However my commitments at work and the commuting I was doing rendered it unrealistic for me to continue my preparations so I wound them down in 2003 and concentrated on my work. I have been doing network design for a number of years and I'm now ramping up to complete my lab prep and do that first CCIE lab attempt in Brussels later this year. I finished my written exam in April so it's full steam ahead now, evenings and weekends preparing.

It's about time!

Uber-Geek

Turgon wrote:

It's about time!

No doubt!

No better time like the present though... I have to start thinking about the #$)*&#ing voice lab soon... *sigh*

boonchin

Well,

17 Oct will be my first attemp in voice lab. So Scott, and how is your preparation in your voice lab?

Thanks and best regards.

Uber-Geek wrote:

Turgon wrote:

It's about time!

No doubt!

No better time like the present though... I have to start thinking about the #$)*&#ing voice lab soon... *sigh*

Paul Boz

Good luck

Turgon

Thanks everyone.

I completed my latest scenario running OSPF and BGP. A few things came up:

If you use tranceivers on AUIs at home dont forget no keepalive if you have no uplink to a switch.

Make sure DCE and clockrate are set up properly on your serial interfaces.

Watch out for typos in your network statements.

Check your ACLs are good for your route-maps so you filter and advertise the right routes.

authentication - dont miss one end of a serial link or it wont come up.

no sync on route-reflector ensuring routes are advertised.

watch out for ip subnet-zero
watch out for ip classless

Take your time with any lab you do. Run routing debugs and check the impact of each change you make. If a router hasn't got the routes it should have, check hop by hop routing tables and the bgp table.

I think if your preparation before lab prep is solid, you have a shot at clearing the actual lab in a 9-12 month window. Start out with one practice lab a week building up to two or three sessions a week. Take your time. Take breaks. On off nights go over your lab notes you will be making with each lab you do. I don't think you need to lose your mind to pass the CCIE. It's a lot of work but mainly it's approach that gets you there. It should not take over your whole life and you can hold down a demanding job and spend time with the family while doing it. Be realistic with your preparation plans. Over the years I have seen many candidates try and 'blitz' their lab preparation to no avail when it comes to getting through. It's too much information. It can affect your pocket, your ability to perform at work and your relationships at home. So be nice to yourself and take your time with your practice labs. Make notes. Reflect. Run debugs.

For preparation before lab prep, do not **** the written exam. Read the books and whitepapers and pass it on merit. You will be pulling on all that reading when you work vendor lab examples. A few weeks before the actual lab exam is a bad time to be unearthing things like deterministic MED. Wendall's Cisco Press book covers this and more. Read it. Get as much exposure to cisco at work as your boss allows. CCNA and CCNP are also helpful.

This weekend I will change the topology at home and do another scenario.

dtlokee

Be realistic with your preparation plans. Over the years I have seen many candidates try and 'blitz' their lab preparation to no avail when it comes to getting through

One of my co-workers just attempted the lab and failed, he had spent 6 of 8 weeks before the lab in training all over the country with one guru or another, spent nearly 40k on it. He would come back and start throwing all sorts of far out there scenarios at me and ask what I would do. Funny part was when he tried to show me the solution, he couln't remember it. I think the best appoach is that of "it's a marathon, not a sprint" and you can't simply cram it all in 3 weeks before your lab attempt.

As for me, I am lost in the pit of multicast dispair.

Turgon

dtlokee wrote:

Be realistic with your preparation plans. Over the years I have seen many candidates try and 'blitz' their lab preparation to no avail when it comes to getting through

One of my co-workers just attempted the lab and failed, he had spent 6 of 8 weeks before the lab in training all over the country with one guru or another, spent nearly 40k on it. He would come back and start throwing all sorts of far out there scenarios at me and ask what I would do. Funny part was when he tried to show me the solution, he couln't remember it. I think the best appoach is that of "it's a marathon, not a sprint" and you can't simply cram it all in 3 weeks before your lab attempt.

As for me, I am lost in the pit of multicast dispair.

I couldn't agree more dtlokee and it's a common problem. There is an emphasis on memorising commands and fixes to certain scenarios that becomes overwhelming in situations like this. Better off baselining what you know about multicasting and concentrating on labbing these topics specifically for a few weeks. Ally that to some serious reading and you should have that topic mostly down. Check the DocCD for references as you configure things. See if you can find some configs for multicast and lab them up, then use show ip mroute et al and post your findings. Try Wendall Odom's Cisco Press book for the CCIE written. Multicast is covered pretty well in there.

Turgon

Saturday 4th Practice Lab.

Today Im doing my 4th lab scenario on my home rack. There is no secret to passing the lab. Regular practice over a very long period of time is the key.

Im working with some old labs while I wait for my IPexpert workbook to arrive. I have removed old content like IPX and concentrated on what is still on task for todays lab.

Time management is important for lab study. This morning I recabled my home rack for my latest lab, configured the switch ports, and erased old configs. Then I went shopping with my wife. Now I'm about to start the exercises and because everything is ready I can simply press on. Mind you, I do have some windows to clean later

This lab covers discard eligibility and HSRP as well as BGP and OSPF. Some NTP is thrown in for fun. Im looking forward to the DE part as I have never needed to work with it in the field. I will post some findings and experiences later once the exercises are done on the 7 router scenario.

For those studying I hope the hours you put in pay off for you!

Turgon

Got the bugs out of the latest lab before calling it a night. It's 11.25 pm here. I will look at the rest of the redistribution situations closely tomorrow. A default route redistributed into OSPF was not appearing on routers connected by Frame Relay. Checked the usual culprits.

Much is learned when you make mistakes and check things with debugs.

debug ip ospf events showed a hello mismatch. This was fixed by amending a bad subnet mask on the serial interface of the router, adjacency formed and the default route then appeared in the routing table as an E1 on the hub router. The further spoke still had no default route though in it's table. Checking revealed a missing network command to put it in area 0. Fixed that and the route is there on both hub and spoke routers now.

I will carry on tomorrow. No BGP planned now, but HSRP, NTP and a RIP/OSPF redistribution scenario with tuning required for optimum routing. A good few hours today on the home rack.

Turgon

About 9 hours all in spread across the weekend on the practice rack involving 7 routers.

Redistribution was tricky with multiple redistribution points, loops and suboptimal routing issues. A constraint in attacking this was the requirement to keep redundant routes in place rendering distribute lists an inappropriate filtering option. Alternative was to use the distance command giving both IGPs (OSPF and RIP) the same AD and then carefully adjust the AD of specific routes leaned on specific interfaces with a better AD. In this way optimal routes are in place and the redundancy in routes remains.

A good deal of time spent checking each routing table and running debug ip routing and debug ip rip to check the routes advertised to routers upstream. One route not appearing as planned was fixed when a bad network statement in OSPF was repaired. Other network statements missing were located and added. Tables were checked with the addition of each command, and resilience checked by shutting down and then enabling interfaces. The routing table was observed during this to see the changes and at times the routes advertised by RIP were examined using debug ip rip. There were no frame relay issues as all subinterface, ospf network type and ospf priority requirements were resolved.

HSRP scenario understood. Unable to do multiple groups with 2500 AUI interfaces. Notes to self include RIP version 2 requirements, RIP authentication key-chains and NTP commands and further study into the redistribution into RIP of default-information originate running in OSPF. Reminder of commands to enter in global mode/interface mode/router igp mode depending on the given situation i.e authentication, DE.

ACLs used for distance commands to define the network requiring changed AD, and DE to define the traffic.

Discard eligibility went well. To test, no need to use debug frame packet, simply telnet to the target IP and run show frame PVC (DLCI) on the upstream spoke, you can see the DE bit counter increment here.

The sun is still shining. I shall take a walk with the family now and enjoy what is left of the afternoon here.

dtlokee

Multicast:

When using MSDP make sure to hard-code your router-id for your routing protocols (OSPF, EIGRP, BGP, ect) since you are basically adding loopback interfaces to multiple routers with the same IP address there is a good chance the router could pick this loopback as the router-id when you reload it (as I found out) and will cause all sorts of problems. Also make sure you either statically configure the RP to use itself with the "ip pim rp-address" command or use auto-rp to announce the loopback as the RP or you will get messages like this on the RP:

Jun 2 12:19:27.155: %PIM-6-INVALID_RP_JOIN: Received (*, 224.0.1.40) Join from 199.1.55.5 for invalid RP 10.10.0.255

other than that I found MSDP to be fairly easy to configure, even using MSDP groups.

HTH

Turgon

dtlokee wrote:

Multicast:

When using MSDP make sure to hard-code your router-id for your routing protocols (OSPF, EIGRP, BGP, ect) since you are basically adding loopback interfaces to multiple routers with the same IP address there is a good chance the router could pick this loopback as the router-id when you reload it (as I found out) and will cause all sorts of problems. Also make sure you either statically configure the RP to use itself with the "ip pim rp-address" command or use auto-rp to announce the loopback as the RP or you will get messages like this on the RP:

Jun 2 12:19:27.155: %PIM-6-INVALID_RP_JOIN: Received (*, 224.0.1.40) Join from 199.1.55.5 for invalid RP 10.10.0.255

other than that I found MSDP to be fairly easy to configure, even using MSDP groups.

HTH

heheh thanks. I'm not sure I need any help on multicast. I think you may have put this in the wrong thread?

Pash

Keep it coming Turgon, 9 hour study on the weekend is very committed, nice one.

Btw excuse my ignorence but how much does frame relay play a part in the CCIE RS track?

Cheers,

dpocoroba

Good blog you have going here. Its definetly worth it now to spend the time to learn things like you are.

Are you doing your own labs or a specific vendor?

To answer that question about frame-relay.

Frame-relay is key to the CCIE R/S lab. Its one of the main points of the lab so its necessary to know it inside and out.


DP

dtlokee

See if you can find some configs for multicast and lab them up, then use show ip mroute et al and post your findings

Just posting my findings from the lab work I had done on multicast for the benefit of anyone reading the thread.

I guess there was more but those bits about MSDP and Anycast RP were the only things I found that didn't seem to be well documented.

Turgon

Pash wrote:

Keep it coming Turgon, 9 hour study on the weekend is very committed, nice one.

Btw excuse my ignorence but how much does frame relay play a part in the CCIE RS track?

Cheers,

Frame always was and still is a vital part of the CCIE lab. Many candidates still fail the lab because of poor frame relay preparation.

This book will help you out immensely.

http://www.amazon.com/Cisco-Certification-Bridges-Routers-Switches/dp/0130903892/ref=sr_1_3/002-1202226-9508043?ie=UTF8&s=books&qid=1180949717&sr=8-3

This book is going for $7 used on Amazon these days. Totally ignored by many of todays candidates it provides an excellent taxonomy of Frame Relay and it's issues relating to DVRPs and OSPF. Buy this book. Even though some of the content its off the lab these days, here are the reasons to buy it.

It's one of the best books ever prepared for CCIE prep and has been around for years.
It covers Frame Relay very well.
The issue spotting methodology used throughout the book is as useful today as it ever was.
Other than Frame it covers other topics like IGPs well.
Old content is still covered well and who knows what you may encounter in the field.

Learning Frame Relay is like crossing the Rubicon for many CCNA's and CCNP's. You can't pass the lab without covering it properly.

Your looking at understanding split Horizon, broadcast issues, subinterfaces, inverse ARP, Designated Router issues, ospf priorities, neighbor statements, frame-maps, ospf network types and LSAs.

The way Frame Relay works with it's logical connections over the physical can introduce acute problems with DVRPs using split horizon, and there are issues with NBMA and OSPF you need to be aware of and counter.

Expect to have to spend a good deal of time reading about these things and labbing them at home with a router configured as a frame switch. In the lab you may be constrained as to what approaches you are allowed to take to get things working properly, so you will need to demonstate a high degree of awareness not only of what problems you have, but which approach is best to use to overcome them.

HTH.

Turgon

dpocoroba wrote:

Good blog you have going here. Its definetly worth it now to spend the time to learn things like you are.

Are you doing your own labs or a specific vendor?

To answer that question about frame-relay.

Frame-relay is key to the CCIE R/S lab. Its one of the main points of the lab so its necessary to know it inside and out.


DP

Im using some old labs I have had lying around for a few years, removing the old content like dlsw. Frame/IGP and BGP is still on topic though. CCO offers many scenarios you can practice at home for each technology. I shall also be labbing examples from Doyle Vol II for BGP and multicast in due course.

I will be using the IPExpert labs any day soon.

Turgon

dtlokee wrote:

Multicast:

When using MSDP make sure to hard-code your router-id for your routing protocols (OSPF, EIGRP, BGP, ect) since you are basically adding loopback interfaces to multiple routers with the same IP address there is a good chance the router could pick this loopback as the router-id when you reload it (as I found out) and will cause all sorts of problems. Also make sure you either statically configure the RP to use itself with the "ip pim rp-address" command or use auto-rp to announce the loopback as the RP or you will get messages like this on the RP:

Jun 2 12:19:27.155: %PIM-6-INVALID_RP_JOIN: Received (*, 224.0.1.40) Join from 199.1.55.5 for invalid RP 10.10.0.255

other than that I found MSDP to be fairly easy to configure, even using MSDP groups.

HTH

Good point, or in IGP design go for loopbacks with an IP higher than your RP loopback IP address and see if that helps you out.

A good intro to finding a routers Rendevous Point using Anycast RP with MSDP can be found on page 726 CCIE Routing and Switching Official Exam Certification Guide Second Edition by Wendall Odom - Cisco Press.

Turgon

Lab no 5.

I have recabled the homerack now to attempt lab number 5.

One the most important parts of lab prep is a good awareness of how to manipulate the routes advertised from one router to another. If your arsenal of approaches to filter routes and inject summary routes is limited the actual lab will kill you. This practice lab looks at some of the options to do such things. Protocols include OSPF, BGP, EIGRP and RIP. The lab also includes some basic NTP tasks and custom queuing, the first QoS task in practice labs.

As a motivator I may increment a counter under my username donating the number of lab hours completed to date. I estimate that including the time required to upgrade my routers, rack mount everything and configure the xyplex terminal server I have so far invested about 48 hours all in on practice labs.

As ever, I keep notes concerning each session I undertake on home and remote racks, and will post findings shortly.

Webmaster

Turgon wrote:

As a motivator I may increment a counter under my username donating the number of lab hours completed to date. I estimate that including the time required to upgrade my routers, rack mount everything and configure the xyplex terminal server I have so far invested about 48 hours all in on practice labs.

Good idea, that'll be an interesting statistic to look back at when you get your number. Is there a more or less official recommend amount of hours for hands-on practice?

Turgon

Webmaster wrote:

Turgon wrote:

As a motivator I may increment a counter under my username donating the number of lab hours completed to date. I estimate that including the time required to upgrade my routers, rack mount everything and configure the xyplex terminal server I have so far invested about 48 hours all in on practice labs.

Good idea, that'll be an interesting statistic to look back at when you get your number. Is there a more or less official recommend amount of hours for hands-on practice?

I have never seen any figures. It's a variable but certainly many hundreds of hours.

It's 1am in the morning here now so Im calling it a night. A problem getting the frame switch working held me up this evening. I have a busy day at work tomorrow. I will carry on with this tomorrow evening.

Pash

Turgon wrote:

Pash wrote:

Keep it coming Turgon, 9 hour study on the weekend is very committed, nice one.

Btw excuse my ignorence but how much does frame relay play a part in the CCIE RS track?

Cheers,

Frame always was and still is a vital part of the CCIE lab. Many candidates still fail the lab because of poor frame relay preparation.

This book will help you out immensely.

http://www.amazon.com/Cisco-Certification-Bridges-Routers-Switches/dp/0130903892/ref=sr_1_3/002-1202226-9508043?ie=UTF8&s=books&qid=1180949717&sr=8-3

This book is going for $7 used on Amazon these days. Totally ignored by many of todays candidates it provides an excellent taxonomy of Frame Relay and it's issues relating to DVRPs and OSPF. Buy this book. Even though some of the content its off the lab these days, here are the reasons to buy it.

It's one of the best books ever prepared for CCIE prep and has been around for years.
It covers Frame Relay very well.
The issue spotting methodology used throughout the book is as useful today as it ever was.
Other than Frame it covers other topics like IGPs well.
Old content is still covered well and who knows what you may encounter in the field.

Learning Frame Relay is like crossing the Rubicon for many CCNA's and CCNP's. You can't pass the lab without covering it properly.

Your looking at understanding split Horizon, broadcast issues, subinterfaces, inverse ARP, Designated Router issues, ospf priorities, neighbor statements, frame-maps, ospf network types and LSAs.

The way Frame Relay works with it's logical connections over the physical can introduce acute problems with DVRPs using split horizon, and there are issues with NBMA and OSPF you need to be aware of and counter.

Expect to have to spend a good deal of time reading about these things and labbing them at home with a router configured as a frame switch. In the lab you may be constrained as to what approaches you are allowed to take to get things working properly, so you will need to demonstate a high degree of awareness not only of what problems you have, but which approach is best to use to overcome them.

HTH.

Thanks Turgon for a fantastic answer. I have heared many people say that grabbing a few 2500 series routers is essential for lab practicing Frame Relay. If you have time would you give some pointers about what equipment is best to simulate frame relay for CCNP and eventually CCIE level.

Cheers,

Turgon

Back in the day, just putting together a very basic lab cost a lot of money. Frame is still on the lab but with the fall in old equipment prices as newer features have required beefier routers you can practice frame for very little money now.

Do yourself a favour. Then spend the rest of this year practicing and asking questions.

Get a 2520 with four serial interfaces, 3 x 2500 series and a few serial cables and you are set. It really shouldn't cost you much. Certainly MUCH less than when I was on ebay looking for a frame switch years ago..what 700 dollars? And that's a four port one. 8 ports cost even more. Even building a tiny rack at home was beyond most folks because of cost. Many people shared equipment. Now, if you use your head, you can assemble a formidable rack at home which is affordable and remote rack the bells and whistles. That said, vendor books require 4 switches so here's where the remote racks usually win and they offer the latest IOS versions. But before you do all that you can learn an immense amount at home with affordable gear and exercises in Cisco Press Books, CCO, free labs and attempt some of the vendor configurations at hoime providing you have the topology.

Buy Bruce Caslow's book the one I mentioned earlier for $7 off Amazon. Do the exercises in that book. It explains how to convert your router into a frame-switch. There are also exercises on CCO.

You need to patiently spend many weeks working with Frame Relay and really understanding it or you will fail the lab. I will be revisiting Caslows scenarios and CCO scenarios myself soon, with plenty of debugs.

On old equipment, get some. It's very cost effective on ebay these days. As the lab content changes more old gear becomes available and it's useful for practice not only for the lab as many fundamentals can still be covered, but also for field work.

Take ISDN as an example. When it was a lab topic, I had to shell out 700 dollars for just one 2503 on ebay a few years ago. I needed two..so that was over 1000 dollars. The isdn BRI switch was another 800 dollars. I couldn't afford that so loaned one for a week.

A couple of years ago, I picked up a four port 2520 (extra frame relay switch, 2 x 2503 routers and an ISDN switch to call my own for less than 700 dollars all in.) Now isdn isn't on the lab, but in the field? You betcha. And you better know something about it when it heads your way at work. And if it does, because of it's age, expect the dial configurations to be ugly. You may be handed a migration job you can't handle if you blew off ISDN because it's 'off the lab'. Get cheap routers with BRI ports and an ISDN switch. Practice with it and keep it handy. More than a few new candidates are shy on older technologies that are still lurking in the real world, requiring support, migration, design. Not just ISDN, but ATM as well. If you can't configure an ATM WAN it's going to expose you to criticism. Im still looking out for a cheap 1010 lightstream when one comes along CatOS as well, if you can stand the noise.

These days with the dash to IP, this protocol has a heavy emphasis on the lab and the IOS is feature rich. We used to learn Appletalk, DECnet, IPX and many SNA bridging variants, DLSW+ and things that were the devil to learn and configure. I would advise you to pick up some older books for reference. Sure, there is plenty of Multiservice convergence going on, but SNA bridging, DEC, VAX, DLSW+ even X.25 is still lurking out there..banks..airlines.

So the flip side of the CCIE rat race is that equipment that is still useful for many remaining aspects of the lab, and certainly for the field becomes available at cheaper prices. So many more people can now afford an actual rack of gear at home and learn frame/IGP/BGP and so much else at their own pace and in their own good time. It took me several years to acquire what I have. Its much more affordable now.

Terminal servers too. They used to be so expensive I was cable swapping my console connection to do labs at home in 2001 until I got the cheaper Xyplex terminal server, which I still use.

I did have 7206VXR's to play with at work though, with the full BGP table, but I didn't own those

dtlokee

If you don't want to buy a 2520 you can use multiple 2501, or even a 2502 as long as it has the IP plus feature set you can create a composite frame switch, a little more complicated but if you don't have a 2520, 2521, 2522, 2523 you could use it.

here is an example:

http://www.internetworkexpert.com/resources/compound.frame-relay.htm

I also like using the 4000 series for a frame switch (they can have up to 12 serial interfaces) but can be found on Ebay for $50-100 with 4-8 serial interfaces.

Pash

Turgon: Thanks very much for your answer. I keep an eye on ebay every day looking for bargins to try and snap up, my training budget is taking shape. I do also have some spare 2500 2600 routers at work to salvage, im gonna have a chat with my line manager and see if i can take some of the redundant stuff home with me for labs, but they want my MS exams done before my CCNP

Keep up the hard work man.

dtlokee: Thanks for the link. That explains a good frame relay setup, i might be able to get some of my workstations at home to fire up dynamips also, if i invest in a little extra RAM here and there.

Cheers,