Ultimate CCIE Rack under $10k?

If you had up to $10k to spend on a rack for home study what would you buy? I'm a CCNP and I want to stay with my security emphasis. Now I'm looking into CCIE but I'm not sure what kind of rack I would need to thouroughly simulate the material.

Goals: Simulate all the technologies of the tests; Keep a security emphasis.

I'm a neophyte to the CCIE track so any help would be ... well, helpful!

Find more posts tagged with

Comments

I'm not quite an expert in this but I know that there is a free CCIE Intro video available on www.cbtnuggets.com which discusses the suggested lab setups.

Check it out, maybe it has some infos for you.

Regards,
Lordy

i would check some of the online racks available to see what they offer and also check cisco site for devices used

Bardamu wrote:

If you had up to $10k to spend on a rack for home study what would you buy? I'm a CCNP and I want to stay with my security emphasis. Now I'm looking into CCIE but I'm not sure what kind of rack I would need to thouroughly simulate the material.

Goals: Simulate all the technologies of the tests; Keep a security emphasis.

I'm a neophyte to the CCIE track so any help would be ... well, helpful!

If I had to do over.....I would invest that 10k in online rack rental and lab books. They keep up with all the new equipment, ios, and lab changes. All you need is an internet connection and a command prompt. A while back, I went out and spent about 8k in a lab setup.....technology moves on, cisco changes there labs, and now I can sell the entire setup for just under 1000 and still have to buy online time.

Kenny

If you go the rack rental route, I've had good experiences with Enablemode.
http://www.enablemode.com/services.html

Marc

http://www.internetworkexpert.com/resources/ccierack.htm

Bardamu. I think the decision is yours to make. Some folks find it easy to do the labs online while others like having the equipment on hand to work with. There are negatives/positives on both ends. My approach has been so far to have some routers/switches on hand and use the online labs to practice on the equip i cant afford to purchase.

Personally speaking, i think i made the right choice on what i did. I would rather have some equipment on hand to play with since there are times when you cant always find available time on an online lab and have to make arrangements ahead of time to practice.

Thats my two cents for what its worth. Hope this helps.

~~ navfett

Nice.

pure hardware for under 10k

2 3550 24 switches, aim for 800-1100 each (one should be emi). 1.5-2k

2 3560 24 switches, aim for 1300-1800 each (one should be emi). 2.5-3k

8-10 2610xm routers with 128/32, aim for 300-500 each. 4-5k

3 nm-4e $350

10 wic 1t $200-250

1 2610 with nm-4a/s and 2 wic2t as stand alone fr switch
or 2 2501 and 1 2520. $150-250
____________________________________________________________
hybrid lab half dyna half hardware.

dual core pc with 2 gig ram and raid 1.2-2k

2 3550 24 switches, aim for 800-1100 each (one should be emi). 1.5-2k

2 3560 24 switches, aim for 1300-1800 each (one should be emi). 2.5-3k

spend whatever left on accessories like nics and xm routers.

if you go with hybrid, expect your pc to look something like mine lol.

Mein Gott!

Is that algae growing in your water cooling plumbing??

DarbyWeaver wrote:

Mein Gott!

Does Dr. Frankenstein know where Frankie went off to?

Hey just kidding...

Nice piece of work.

But may recommend also that there are Quad Nics (4 per PCI Card) for about $25-100.00 on ebay as well.

Just in case.

the quads were more exp, and harder to find. If you can send me links of your source for the quads please do.

TheShadow wrote:

Is that algae growing in your water cooling plumbing??

even with distilled, over time it will get like that. I should switch to antifreeze.

heres a benchmark for the dynawhores like myself...
If you have suggestions please comment.

2 pcs (one dualcore, one single)
3 gig total ram
1 real router
1 real switch

i guess i should try with pure virtual mem, should be slower though or no?

mmap set to true with an extra router......

awfully quiet in here so i had to post another dynashot lol

Thanks

slightly slower than normal for me, but those screen shots are barely configured, I'd imagine that if all the routers were running a few routing protocols, it would get slow or even crap out. Normally i run only r1 r2 r3 r4 r5 r6 with one frs and 3 real bb routers and during the start of a practice run utilization is close to nothing but towards the end it will reach as high as 3/4 and higher sometimes with just 6 main routers.

pc1 main specs

2.5 socket 775 celeron (grr)
1 gig ddr400 corsair
raptors in raid

pc2 main specs

2.4 dual core amd
2 gig ddr400 gskill
raptors in raid

Pc2 basically handles 2-2.5 times as much as pc1. With a new core 2 duo setup id say about 3-3.5 times or more than the pc1.

here is screenshot of normal dynamips usage but with only 6 main 128 dram routers. IP addressing and frame relay connectivity are the only things that are applied to the 6 routers. Once i apply routing protocols and more in depth configurations, utilization gradually increases in 3-5 percent increments. With more routers running there is a higher starting utilization and the 3-5 becomes 5-10 and higher increments. I barely started using uncompressed images few days ago- the performance and startup time is much faster now, startup use to take about 30 seconds with compressed, now 5. So for whoever that havent or doesnt uncompressed their images yet, do it (winrar). Anyone with comments or suggestions please reply.

What are the specs of the machine is this running on? I have heard of people in the hacki forum using threadmaster to help with performance issues, but have no experience with it myself.

Also using one window and reverse telneting rather than opening a window per router would probably help a lot, I know it does on my machine.

pc1 main specs

2.5 socket 775 celeron (grr)
1 gig ddr400 corsair
raptors in raid

pc2 main specs

2.4 dual core amd
2 gig ddr400 gskill
raptors in raid

Pc2 basically handles 2-2.5 times as much as pc1. With a new core 2 duo setup id say about 3-3.5 times or more than the pc1.

thanks for the comment, Ill give it a try, i like the multiple window feel more than the access server feeling for some reason.

apd123 wrote:

Also using one window and reverse telneting rather than opening a window per router would probably help a lot, I know it does on my machine.

multiple windows also useful for debug purposes.

And now let the games begin

And here is one last bench lol... promise
http://home.comcast.net/~noshut/last.gif
now just need about 15 more nic lol, jk
*back to the lab cave*

the above is from an old amd x2. Imagine what 2 new core 2 duo or quad, or even itx cluster could do!!! 100+ routerz lol=)

you must remember, no matter how fast your PC if you are running a boat load of routers it will peg your CPU at 100%. When its stuck at 100% its slow as sh!t to type commands and you will often lose your protocol adjacencies.

HELLZxPHADER - nice work. Does the setup work for realistic labs? With those pictures of switches I assume you are trying to emulate the current CCIE lab environment? Just curious as I have though of getting a couple of extra NICs for my PC and buying a couple of switches. I have only tinkered with Dynamips / Dynagen. I have got it up and running but have not done any lab work on it (yet).

So when you fired up frakienPC up for the first time did the lights in the neighborhood flash, and did somebody yell "its alive!!!"

I assume you have no cats, dogs, or small children in the house otherwise you would probably cover that bad boy.

ModemHumper wrote:

you must remember, no matter how fast your PC if you are running a boat load of routers it will peg your CPU at 100%. When its stuck at 100% its slow as sh!t to type commands and you will often lose your protocol adjacencies.

Please remember that when i try to run 10 or more 12.4 routers from one pc, the main purpose is benchmarking and not intended for lab simulation study.

And you are right about the slow response times as a result of the number routers running, but in my opinion "slow" only occurs on single core cpu while dual core pc will run slightly slower than normal.

mtbcyclist wrote:

HELLZxPHADER - nice work. Does the setup work for realistic labs? With those pictures of switches I assume you are trying to emulate the current CCIE lab environment? Just curious as I have though of getting a couple of extra NICs for my PC and buying a couple of switches. I have only tinkered with Dynamips / Dynagen. I have got it up and running but have not done any lab work on it (yet).

So when you fired up frakienPC up for the first time did the lights in the neighborhood flash, and did somebody yell "its alive!!!" I assume you have no cats, dogs, or small children in the house otherwise you would probably cover that bad boy.

Thanks for the comment, here is break down of how I run a simulated r&s lab...

This diagram only displays the physical and simulated physical layout of a default lab

http://home.comcast.net/~noshut/PHYSTOPin.jpg

Depending on the type of lab running, usually i will only have to move around eth cords, I try to not change any dyna configurations.

Believe it or not... when i run a normal dyna lab, utilization remains low untill the end of the lab and even towards the end the load is still 1/3 from being full (it uses less power than say bf2)

Quick Links