The Hardest Phone Interview in the World!!

wildfire

Well today I had a phone interview with a CCIE for a network consultant role, Anyway It lasted well over an hour on the phone and it was some real hardcore questions!

Examples were,

Can you tell me the fixed part of a HSRP MAC address?

Convert 224.001.123.123 into a MAC address?

Can you represent 8.5Mb as an OSPF cost?

What is a type 9, 10 and 11 OSPF LSA



Bear in mind this is all on the phone, Ive had had a few technical interviews, this made the CCIE written look like childs play !

They called me to say that they were conrcered that my technical level wasnt as high as they thought but would like to see me face to face, do you think I should bother? Its a once in a lifetime job, but I felt very aukward on the phone, and Im sorry I dont know was said a few too many times!!

Anyway was wondering has anyone had anything so intense before?


EDIT

Went to Interview, job offered, job accepted

garv221

Wow, pretty intense. Sounds like a CCIE being a stiff. I would go in and give it a shot face to face. What does the position detail?

Olajuwon

You should expect those types of interviews with your level of Cisco expertise. I am no Cisco expert, let's wait for mikej412's input on those questions. Anyway, you should give the face-to-face interview a shot. You have nothing to lose by going, except for a loss of confidence if things don't work out; ehehehehh Good luck!

rockstar81

I am no Cisco expert, in fact I have only just started studying for CCNA however during one of my courses with a CCIE instructor (bryantadvantage) he made a good point which I picked up on, he said the ? Tool is excellent to use and that it was impossible for any Cisco expert out there to remember every switch and every command so use it. I picked up on this as in my naivety I thought a CCIE would be some oracle of the computing world but it just goes to show we are all human and have limits. I would go to interview anyway and think you could be pleasantly surprised with the outcome.

wildfire

Your right I will go to the interview an face it out, dont get me wrong I expect hard questions, Ive had 3 phone interviews this week, most of them start easy and lead up, this one was very hard from the word go, things like the Mac address conversion, I can do easily but on the phone on the spot its not so easy trust me!

Those were just examples there were some harder questions about route dampening and MPLS TCs that I cant even remember! Even CCIEs use the DocCD you know! thats why they give you access to it in the lab.

Heres one question for Mike that I remeber,

How do you control a Multicast over a hub and spoke frame-relay link in such a manner that doesnt flood all the links simultanously when the broadcast keyword has been added?

This one even with a bit of post research I cant answer.

Any thanks for the input guys

garv221

wildfire wrote:

How do you control a Multicast over a hub and spoke frame-relay link in such a manner that doesnt flood all the links simultanously when the broadcast keyword has been added?

This one even with a bit of post research I cant answer.

Any thanks for the input guys

What about stub IP multicast routing.

Here is a pretty cool lab I have in my favorites. It has alot on Frame Relay with lab examples. Hope this helps.

http://www.internetworkexpert.com/downloads/iewb-rs.sample.lab.pdf#search=%22CCIE%20spoke%20and%20hub%20frame%20relay%20multicast%20flooding%22

EdTheLad

wildfire wrote:

Heres one question for Mike that I remeber,

How do you control a Multicast over a hub and spoke frame-relay link in such a manner that doesnt flood all the links simultanously when the broadcast keyword has been added?

This one even with a bit of post research I cant answer.

Any thanks for the input guys

On a multipoint interface you can use Pim-nbma , if you have the physical interface divided into multiple ptp subinterfaces i' think all you need to do is
use pim spare mode, set the hub to be the RP and multicast traffic will only be sent on the subinterface that receives a request.

Humper

Go to the interview face to face with this CCIE. Ask him some of the most ridiculous questions possible. Like a vulnerability in some IOS....Go there to piss him off Talk about being a dick

Don_PA

Hang in there wildfire.

I had off the wall phone tech interviews in my life. I prefer a face to face interview. Let's face there are alot of good tech guys who know their technical knowledge from the books, education and hands on experience but do not have the oral communications skills to effectively explain and communicate that knowledge in clear manner. Nevertheless, it is good to practice either alone in with your delivery of explaining technical terms like Kerberos, EAP, OSPF, BGP, VPN, NAT, DHCP, DNS, route summarization, War Driving, EIGRP, Frame relays, and such.

I couldn't explain those tough questions like that over the phone, but you should feel better prepared in an actual face to face interview when they(the interviewer) will also judge you by your apperance and delivery in the interview. Don't cave in or suck up to these faceless interviews and if they didn't like your response , the heck with it and move on to another potential employer who may want to interview you face to face and judge you not only by your technical saavy and skills.

wildfire

On a multipoint interface you can use Pim-nbma , if you have the physical interface divided into multiple ptp subinterfaces i' think all you need to do is
use pim spare mode, set the hub to be the RP and multicast traffic will only be sent on the subinterface that receives a request.

While I dont want this to into a pure technical thread, The broadcast keyworkd sends 1 multicast across multiple links on a P2MP its my understanding that PIM does not prune the path back in that manner on a frame cloud. Also I think what he meant was make it go over all the p2mp paths from the hub but NOT all at once! else it would exceed the CIR/BE.

EdTheLad

wildfire wrote:

On a multipoint interface you can use Pim-nbma , if you have the physical interface divided into multiple ptp subinterfaces i' think all you need to do is
use pim spare mode, set the hub to be the RP and multicast traffic will only be sent on the subinterface that receives a request.

While I dont want this to into a pure technical thread, The broadcast keyworkd sends 1 multicast across multiple links on a P2MP its my understanding that PIM does not prune the path back in that manner on a frame cloud. Also I think what he meant was make it go over all the p2mp paths from the hub but NOT all at once! else it would exceed the CIR/BE.

It's too late , this thread is offically gone technical.
The broadcast keyword will allow mulicast traffic to flow on the pvc.If you were using dense mode pim it would flood the mulicast on all pvc's.If you are using spare mode it will only send a multicast stream when a request is received.If you have multiple subinterfaces each on its own subnet, the subinterface will be able to work like a port and send multicast streams on a single pvc when a igmp request is received on the particular subinterface.Controlling the traffic so it doesnt exceed CIR doesnt seem feasible.If you run multicast video the end device would need a hugh buffer to take the multiplexing into account,maybe its possible but i havent heard of this.
The pim nbma can control multicast to the extent that it can choose which neighbors to send multicast traffic to,therefore you wont get a flood on all pvc's on a multicast interface.
Anyway this is my idea, maybe Mike can confirm.

mikej412

wildfire wrote:

do you think I should bother? Its a once in a lifetime job

Go! Even if you don't get it, you'll have an idea of what to prepare for when the next once in a lifetime job shows up! Take the shot!

wildfire wrote:

Anyway was wondering has anyone had anything so intense before?

Not on a job interview.... but a 2-1/2 hour techie showdown across a boardroom table to save a multi-million dollar deal. And then there was the 2 day techie pi**ing contest that I broke up with the real answer.

Consulting positions usually means one of two things -- someone who looks good in a suit and can rack up billable hours -- or someone who can go in and get the job done. This one sounds like they are looking for the 2nd one.

The technical level thing could be just that -- they thought you were farther along in Lab prep than you are... or expected you to remember more trivia from written exam study. They still invited you to the dance.

Olajuwon wrote:

let's wait for mikej412's input on those questions.

Okay -- phone interview, huh? Sounds like a job for a headset and a quiet keyboard!

wildfire wrote:

Can you tell me the fixed part of a HSRP MAC address?

Convert 224.001.123.123 into a MAC address?

Can you represent 8.5Mb as an OSPF cost?

What is a type 9, 10 and 11 OSPF LSA

An hour? Do you know this CCIE from somewhere? Did you ever get him mad at you?


0000.0c07.AC & then the group number


0100.5E and the lower order 23 bits of the IP


10^8/bandwidth in bps......
convert 8.5Mb.....
uh, whats the reference bandwidth?


Is Cisco supporting those.... I'd have to go to the RFC

is right.

Interesting. Technically, with those quick things I'd blurt out -- that would be 25% -- since I didn't actually come up with the MAC address, or give a number for the OSPF one.... and my "go to the RFC" answer would have included "rolling my eyes" at the question if it had been face to face.

But I might have gotten partial credit (or even full credit)..... and if the interviewer wanted the "real answer" -- I'd have grabbed pencil and paper and talked through it (and cursing under my breath).

But a lot of times in technical interviews, its the thought process that goes into the answer that's more important than the answer.


The HSRP I remember because that was one of the things I figured could be asked on the BCMSN exam..... and I got hammered by HSRP questions on the real exam -- so I re-read all the HSRP & VRRP stuff after that.... Plus its on already on my ports/macs to remember note sheet I'm putting together as I go back through everything for the CCIE Lab.

The convert the multicast to IP -- I finished my first refresh on multicast this week -- that's an easy one. I did hit lots of other stuff that I never had worried about remembering -- so the next pass will be just to write down what I need to try and work through hands on. I glad I did add a Multicast book to my R&S reading list.

The OSPF was covered pretty darn well in the Network Academy Course.... but it gets annoying (and confusing) with the 100Mb reference.... then the change to 1Gb... and now you have to worry about 10Gb....

Yeah -- I'd stick with the standard 10^8/bw in bps since it is a "slow speed question" -- but slip in the issue with the reference bandwidth in case of speeds 1Gb or higher.


The LSA 9, 10, 11.... um, I read the RFCs years ago.... and after about 2200 I didn't need to keep up to date, and after 2800 I stopped maintaining my full text indexed database. The only thing I remember about these was there was nothing worth knowing (this was in my UNIX networking days). Maybe its something Cisco is going to use to support their new "Converged Network" focus -- but I got my LSA notecard and it doesn't go higher than 7. I can't decided whether to worry about it or not.

ed_the_lad wrote:

Anyway this is my idea, maybe Mike can confirm.

Yep, I think you had it at PIM NBMA No auto-RP... And ask for clarification if there are only Cisco routers, or some non-Cisco routers involved if they wanted more than the PIM NBMA.

If it wasn't hub and spoke (multipoint), then for p2p I'd have gone PIM sparse mode. Still no auto-RP.

wildfire

Thanks mike you BRAINIAC!! next time you can take them for me, lol

Well I went in for the interview.... 3 hours later I came out phhheeeww.

Generally felt good, but 3 hours I have never had an interview of that length before! and he said this was only the first round of interviews! damm! A load of HR type Qs, followed by another intense technical test, which I was better prepared for, apart from falling down on IP protocol numbers, I got 89 (OSPF) Right, MPLS-IN-IP wrong , IPv6 wrong . Also was a little weak on the IPsec stuff, so I need to hit the security books!
Next phase sit down at a bunch of equipment and do a series, most easy, config etherchannel, only allow telnet from host x, vtp, dtp, RSTP, bit of pim and diffserv. I cant wait for round 2 (hopefully)

sprkymrk

Good luck!!!

Olajuwon

Wow, that's intense. Best of luck, dude.

TeKniques

I hate threads with the long link that I have to scroll across the screen to read!!!!!!!!!

Anyways ...

wildfire wrote:

Thanks mike you BRAINIAC!! next time you can take them for me, lol

Well I went in for the interview.... 3 hours later I came out phhheeeww.

Generally felt good, but 3 hours I have never had an interview of that length before! and he said this was only the first round of interviews! damm! A load of HR type Qs, followed by another intense technical test, which I was better prepared for, apart from falling down on IP protocol numbers, I got 89 (OSPF) Right, MPLS-IN-IP wrong , IPv6 wrong . Also was a little weak on the IPsec stuff, so I need to hit the security books!
Next phase sit down at a bunch of equipment and do a series, most easy, config etherchannel, only allow telnet from host x, vtp, dtp, RSTP, bit of pim and diffserv. I cant wait for round 2 (hopefully)

Holy crap! Best of luck to you. This shows just how hard some of these jobs are to get!

wildfire

Well out of this I got offered 2 jobs, one is is 3rd/4th line operation support, the second was the origional role, the design and build consultancy, the 3/4th line actually pays more (£3.5K) because it includes a callout allowance. But the second role is better technically.

Do I take the slightly better role for less money? my main motivation is which has the best lead up for the CCIE lab, support or design and build outs?

sprkymrk

You gotta do whichever one will keep you motivated and interested. Especially if the pay is not too much different. I think it's about $6500US correct? I guess it depends on how much you need the money.
Good luck and congrats!

wildfire

thats about right conversion wise, I dont NEED the money, but money is always nice, but I have to do what my heart tells me I guess, Im leaning to the consultant/design role.

Olajuwon

Go get'em. Cisco engineers make more than MS engineers for a reason. CLI

uke

sprkymrk wrote:

You gotta do whichever one will keep you motivated and interested. Especially if the pay is not too much different. I think it's about $6500US correct? I guess it depends on how much you need the money.
Good luck and congrats!

3500 pounds=$6500 now? yeesh! has the dollar gotten THAT bad or is the pound just THAT strong now (or a combo of both)?

D-boy

The pound is just strong...

I used to live in the UK, London, and moved to the US in 2002...

Now when I go on vaction I have to save double the amount to even have decent spending money... but when I lived in the UK, I could go over and shop like a king, as the exchange rate was always good...

It would be hard for me to move back to the UK, as I would have to save a ton of money to have enough to get by for the first weeks months, along with trying to find a job...

wildfire

yeah a combo of both, 5 years ago the exchange rate was about $1.4 = £1 now as a combo of a weak $ and strong £ its about 1.9/1 which means the wife wants shopping trips to New York more often!!

strauchr

Olajuwon wrote:

Go get'em. Cisco engineers make more than MS engineers for a reason. CLI

They do?

I use the CLI quite a bit. Maybe we should get paid per word we type in our jobs because that proves everything

Olajuwon

strauchr wrote:

They do?

I use the CLI quite a bit. Maybe we should get paid per word we type in our jobs because that proves everything

Relax, dude. I was being sarcastic above.

strauchr

I was just having a dig

darkuser

hmm i know those things but would not have been prepared to do it
off the top of my head ....

looked this up .....

Q. What does the HSRP MAC address look like?

A. The HSRP MAC address is derived from the group number, and looks like this:

0000.0c07.acXX
XX is the HSRP group number.


Q. I used to be able to configure more than 16 HSRP groups on my Catalyst 3550. I upgraded the switch from Cisco IOS® Software Release 12.1(11)EA1 to 12.1(13)EA1, and now I can not. What has changed?

A. If you have more than 16 HSRP groups, you could trigger connectivity issues. Therefore, such a configuration is not possible as of Cisco IOS® Software Release 12.1(12c)EA1.

i would go see them if they want to see you ....
do you think they want to waste their time ?

to not go would be like not taking your at bat
(you cant get a hit if you don't swing)

many actors will tell you of how many calls they went on before they hit it big. the ones that stopped going are the ones you've never heard of ....

my 2c

mikej412

wildfire wrote:

support or design and build outs?

High level support can be fun sometimes -- like when you only get to deal with the top 1% of the problems -- the ones no one else has been able to solve. Oh wait -- I was doing design and build out stuff at the time -- nevermind. No matter what you are doing, if people think you're good, they will hunt you down for help on the "fun problems."

The Design & Build Out sounds more interesting on a day to day basis. If we're voting -- I cast my vote for Design & Build Out!

Congratulations! I guess you took the shot and hit two targets at once!

wildfire

Cheers mike,

I opted for the Design and Build role, the being on call bit is a bear.with the support! back to permenant employment after contracting should be intresting, but my current contract has plenty of time but no fun or steam left in it so Im bailing at the right time, although Ill be on less $$$ but hey
No I just the need the time to read and work through this lab book I splashed out on.