The collected works of WolvesOfTheNight - Page 1

Sometime kindof soon I will be needing to get a real job, and I decided that having a cisco certification would be useful. I don't really think that they teach you all that much in the certification, but it looks like it makes a big difference when a prospective employer considers you.

I got the 2 self study books (INTRO & ICND) and have been going through them. It seems to me that they have a huge focus on memorization, but maybe that is just be. I tend to think that memorizing things is a waste of time - what you don't know you can look up, and if you use it often you will quickly learn it because you are using it. How many people have a boss tell them "do this, and you can't look up any information?" Anyway, that is beside the point.

What I am really wanting to know is a) how close are the practice tests to the real thing; b) about what grade is passing (90%? 80%? 70%? 107%?), and is there anything specific that I should worry about?

If you have anything helpful to say (without violating cisco's test agreement thing) I would appreciate it.

Well, I am going to go back to wishing that IOS had a halfway decent shell.

OK, that is about what I was expecting. I will spend more time on subnetting, though I am OK at it right now (my big problem is making little math errors). Actually, the binary math is something that my CS profs moaned about in a number of classes and I am OK at. Also, the subnetting seems more like a more useful thing to know than which trunking protocols fully encapsulate the data and which just add a header (one of the practice questions asks something like this).

Otherwise, I will go through your list and see how I am doing on each thing.

As for the shell thing, I don't thing that it my terminal setup (same one that I use on my O200). Mostly, the IOS CLI seems like something that I would expect out of the 70s or 80s. It has a very basic help system and very simple interface. For example, you can not edit an access control list - you can append it or delete it. My book recommends that you type it all up in a text editor and then paste it in. Come on, would it really be that hard to include something like vi or emacs (or even pico) to edit such things?

Back then it was really important to keep the load on the processors down, and a spiffy shell could take up a good bit of power, this made since. Now we have faster computers and can use better shells (or in many cases waste them on iCandy).

Yes, I know that there are windows programs that provide an easier interface to set stuff up. You can save the config file elsewhere and edit that (though the CCNA book has not said much about actually doing this). The CCNA stuff wants you to use the IOS CLI. Also, the windows programs won't work on linux, mac, or my O200 . Well, OK, I would not drag an O200 around to talk to a switch, but you get the idea.

Anyway, I will let you know if I have any specific questions. Thanks!

That is worth knowing. My CCNA book just said that you can't do it! If I end up with a job doing a lot of work on cisco stuff I expect that I will need to learn a lot more about them. Just working with the stuff is the best way to learn it. The problem is that telling people in a job interview that you don't know it but can learn it does not sound nearly as good as being able to tell them that you know something about it. Thanks!

My books (2005 edu) admit that the test does not use 1900 switches! They say that you do not need to know them for the test, but have an appendix on them in case you have some around.

Good luck on the CCNP! I doubt that I will ever do one unless my employer wants me to. I feel that just being able to do stuff is far more important than the certification, but I hear that managers do not always see it that way.

Are the simlits something like you have 3 routers connected to each other, a switch connected to each router, and 3 computers on each switch. Then they want you to subnet a class A network, assign IPs to everything, and make it all work? There was a practice section in the book like this.

So far, in the practice questions the worst that I have had to do is configure 2 routers in some simple way, which did not take very long.

Hamei - you have a good point, and it is one big reason that I am working on the cert (in an interview that I had awhile back they liked the fact that I was working on one). However, I really wish that there was a better way to do it. I really don't like the feeling that the companies are just selling certs to make more money more than to help people.

Ipaddict - thanks for the info. I have had mini problems in the test questions, but not any big ones. Hopefully I will be ready, though you are right about the best way to learn being through real life experience. Of course, that makes the chicken & egg problem of needing a job to learn the stuff for the cert and wanting the cert to help get a job in the first place...

Yes, it does not look good if you only have a cisco cert. Well, I have a BS CS from NM Tech. The main point of getting the cert is to a) learn some more about networking, and b) make me stand out a little more when applying for a job. Despite occasional thoughts of leaving the nation, I will be looking for a job in the U.S., probably somewhere in Colorado but outside of the cities. I know, if you don't do the Denver corder it is hard to find much there, but I will try. So don't need to worry about my coming to you with just a cisco cert. However, if I hear of anyone your area with just a cisco cert I will send them your way .



Anyway, I have another question: what is the best way to get some practice? I have been messing with the books & the netsim LE software that came with the books. However, I really need to do more than just read the book & have netsim LE tell me to upgrade to the full version when I try to do something. Some people seem to be able to get it by just reading the book, but I am not one of them. I need to do a lot more real stuff to really learn it. I can do a job given the book and some time, but they don't let you use the book on the exam...

After deciding that the book was not enough I checked into taking a class. The only place that I found near Socorro (in this case, in Albuquerque) offers a typical prerecorded class & simulated lab setup, with a T.A. They have a few real routers to goof with, but not enough to do full labs on real hardware. The class is 10 3.5 hour sections. There setup would probably work, except for the price; they want $2595 for it. Discounting room & board, a semester at NM Tech costs less...

This leaves me with the conclusion that I am probably best off buying some simulation software. I don't like the Boson NetSim, LE, though many of my problems are due to the fact that it is always trying to get me to upgrade. Also, I use a Mac at home for my primary computer, and have to run the thing under virtual PC, which is annoying. It is tempting not to buy anything from them because of this, but getting the certification done is more important.

What would you recommend for practice? Should I buy the full version of the Boson NetSim, or something else? Is there a better way to do it?

Well, looks like the full Boson CCNA pack is about $300:

http://www.boson.com/Product/CIS-KK-CCNA-01.html

It is way more than I want to pay, especially given that it is wintel only and they don't even bother to say so on their webpage. On the other hand, it is almost as much as the $2500 class offers and may enough to learn the stuff.

Anyone used this stuff? Is it actually worth buying?

It depends on how you look at it. For me, it will not really make much difference on the job. Everything that it covers is stuff that I would easily look up if I ever needed it on the job. For that matter, I disagree with some of the ways that they do things, and will just have to know the "right" answer for the test. The ability to think for yourself & research the stuff you are working is far more important than any certification. I wonder how many people set the enable password to "cisco" just because that is what they always did in the CCNA study...

Also, as you may gather from my previous comments, I am not all that impressed with cisco. They make some ok stuff, but I don't think that they are that good of a company. They stay in business because they a) are the largest networking company out there, and b) they keep buying out their competition. If I get a job where I decide what switches to purchase there is a good chance that I would not get cisco. They cost more, and most people don't need the features that are unique to cisco anyway. Besides, if I designed a network, I would avoid proprietary software, hardware, and protocols, as much reasonably possible in order to avoid vender lock in

However, something is also worth whatever someone will pay for it. In the U.S. there seems to be a lot of desire for certifications. I think that the certification thing is taken way too far, but I can't change it (yet). As previously noted here, some people make hiring decisions based partly on CCNA certifications. Some places will also pay you a bit more if you have one. During layoffs they will give you a slightly better chance of not loosing your job. Do I think that this is the best way of doing things? No. Do I think that it will pay for me to get a certification? Yes.

I think that there is a good chance the certification will help me get a job, and possibly get me a hire pay. Thats why I am getting it. Sadly it is not really about learning stuff (though I try to learn as much as I can from it), but about how the U.S. job market works.

As for a CCNP, well, I will worry about that after getting a job. If my employer likes the idea I probably will; if they don't care I probably won't.

Well, the boson software has some annoying requirements:

-M$ windows only
-It uses the internet to do a large portion of stuff (they would not tell me how much)

The windows part is expected (I really wish they would make it for Linux or Mac or Irix or something else). However, the network part is a big problem because:

-I don't have any antivirus software for my windows stuff (I never connect it to the net).

-I am stuck with an ultraslow dialup connection; I am lucky to get 28.8kbps.

-Being a dialup connection I don't want to tie up the phone all the time.

Anyway, I looked around some more and am going to see if the Semsim software will work. It costs less anyway. Anyone around here used it?

Actually a bit under 3 years. And, yes, there were some very difficult managers. I got tired of having to worry about irrational reactions from a manager from outside of my chain of command.

On the other hand, I can't really put on my resume that I did my boss' job while I was a student worker. I suspect that most prospective employers would think "Yeah, and you also invented TCP/IP while you were at it; next candidate." Thankfully, if someone contacts my references they can at least find out that I did far more than a student was expected to. Overall I would say that the various jobs I had as a student (the computer center & LANL) gave me a huge amount of experience beyond just having a CS degree.

While I have not had a real (i.e. 40 hours / week for more than a summer) job, I get the feeling that there are a lot of incompetent people & companies out there. I always find it a little scary when people complain that dilbert is too realistic. Hopefully I can get a job without any major troublemakers, but I doubt that I will. As is I am pushing it by trying to get a job in Colorado and not in a city (take one look at the colorado listings and they are spamed by jobs in the Denver metro area and CO springs). Well, maybe I will get lucky...

A quote from one of the boson labs: "...this is a trick question. This is the kind of misdirection you might see on the exam." Sigh... I sure hope that they are not really into trick questions. They tend to get me, and they are also one dumb way to test networking skill; they really test your ability to detect trick questions.

Oh, and while I am not fond* of the boson netsim, it appears to be good enough to learn what you need for the CCNA exam, and the price was right (thanks, ipaddict).

Well, I hope to be ready to take the exam soon, and then I can worry more about finding a real job.

e.g: It is windows only, and crashed half way through a big lab. Grrr...

What happens if you don't finish in time? I assumed that you just miss the questions that you don't have time for, like on most timed tests, but I thought I should ask. Anyway, I am fairly good at subneting calculations, but I will practice more to be safe. I wish they would let me use my HP48gx calculator on the exam. It will do all the decimal to bin to hex to octal conversions; it will even do bitwise operations, such as AND & OR!

They still cost a lot used, despite having been first released in 1993! I am not letting go of mine, and I am keeping an eye out for a spare (thankfully these things seem to last forever, unless they go swimming). I also have an HP 49G, which I hate; they made it dual algebraic / RPN, and in doing so sacked some of the usability of the RPN. They also used some nasty rubber buttons. I started on a 48G, it went for a swim, I got the 49G, hated it, and then got a 48gx. Maybe I will try to sell the 49, or trade it for another 48gx...

Quick question: how brain dead is the grading of the test? Working with the boson software, it does not always seem to recognize that there is more than one right answer. For example, using sub-interfaces s0.1, s0.2 &, s0.3 vs s0.2, s0.3, & s0.4 when setting up a frame relay network, or that there are many ways of setting up an ACL that does what you want. Also, if you have a large simulated set up problem (e.g 4 routers & 4 switches), and get one number wrong (e.g. mask 24 instead of 25), do you get the whole problem wrong, like you do in the boson testsim?

Do they have an intelligent way to grade the real tests, or do you have to worry about the computer program marking you wrong because you did not happen to write an identical ACL to the one they wrote? I am guessing that they would have some way of properly grading them, but I don't credit cisco with copious quantities of intelligence.

Well, I took the test today and passed it! Their simulated router tests are not as complex as expected, though I was unsure if I was supposed to do a copy running-config startup-config on a couple of the questions.

The cisco test program decided that my good understanding of planning and designing outweighed my lack in implementation and operation. I guess that if I get a job with cisco hardware my employer will just have to put up with my checking a book or webpage for the syntax of a few of the commands. If I get an evil manager that refuses to let me use documentation then I am doomed, but I should otherwise be fine. Besides, I doubt that there are very many real life situations where you are given 5 variations of the same command and have to choose the one with the correct syntax.

Now I need to try to get a real job. I want one in a town in the mountains of Colorado (i.e. NOT the Denver / CO springs / FT Collins area). I fear that this may be asking too much, but it is worth a try.

Humm... While I have sometimes considered leaving the US, I don't think that China would be the best place for me. I doubt that their government would like me, and vice versa. Also, I am terrible at learning new languages, and the difference between the oriental l languages and English makes them especially hard.

Out of curiosity, would jobs in China actually compete with jobs in the U.S.? In the U.S., the only time you really hear about jobs in China is when some large cooperation is accused of employing sweat shop labor. Also, those stories tend to be lacking in real information.

I don't know if I will do a CCDP or not; it depends on what job I end up with. If my manager wants me to I will. If they don't I probably will not.

I picked up a CCDA self study book thinking that it would be more interested in it than memorizing IOS commands. Glancing at the first chapter I almost ran away screaming due to a buzz-word overflow. For example, here is a partial list of the buzz-words that show up in the chapter one 10 question "Do I Know This Already" quiz:

customer focus (5 times)
key constituencies (4 times)
leverages (1 time)
competitive advantage (1 times)
economical constraints (1 times)
decentralization (2 times)
centralization (1 time)
continuous standardization (2 times)
core versus context (5 times)
core versus edge (2 times)
market parity (3 times)

I will look at the next few chapters hoping that since the buzz-words lack substance they all floated to the top of the book, leaving the content in the rest of the of the book. If they have some really good information later in the book I may go through it and then take the CCDA test. However, I want sold information on how networks are designed, not a pile of buzz-words large enough to obfuscate the US national debt.

The other thing I wonder about with putting huge amounts of effort into cisco certifications is that I am unsure if they will hold their value for another 20 years. It seems to me that cisco was really on top of things 10 years ago, but they seem to have stagnated. On the other hand, maybe I am just biased because they required me to spend lots of time studying ISDN, which I think was one massive waste of time. How many people do you know that are installing ISDN on new routers so that they can use it as a back up network connection?

DraconianTimes - I agree with you about the importance of knowing a wide range of products. I generally try to keep any complex setup as vender independent as possible. It amazes me how much some people will put up with vender lock in.

I wish that they would either give up on ISDN or make it something worthwhile. It would be cool if you could have a ISDN / DSL cross where you had 50 64KBps channels for the cost of 3Mbps DSL. Then you could have 5 people talking on the phone and surf the net at the same time. You could also dial half your network connection to one place and the other half somewhere else... However, as ISDN is now (speed & price) I think that they should just give up on it.



I don't have any real experience with the large scale networks, so you may well be right. I really just wondering if the issues that I see with cisco are part of a larger problem or not. My observation is made from a very limited prospective, but I can give a few examples of where I think that they are behind the times.

-As previously discussed, they seem to be big on ISDN. Despite ISDN having been out for about 20 years it has not caught on. If cisco is going to push ISDN for computer networking they need to make it compete with modern technology, both for cost and bandwidth. Otherwise, they should stop wasting our time and their money and let it die.

-One of the pervious big reasons to buy cisco was that they were multiprotocol. Now most everyone is happy with IP; who cares if your router will handle appletalk or not?

-For ethernet routers/firewalls/multilayer switches you can put a pile of ethernet cards in a linux box. I don't know how this really compares with using cisco products, but I do know of people that do it because it is cheaper. Actually, I often hear of students doing it in their dorm room in order to get around network regulations. I may do it with my O200 for the fun of it. What happens when those guys get real jobs? They also say that it does a better job & is easier to setup, but I don't have the experience to comment on that. However, I do like the vendor independence of such a solution.

-Judging from what I covered in the CCNA certification, they don't seem interested in making their stuff easier to administrate. While this makes it easier for me to get a job, it is a poor long term business practice.

Anyway, I could be wrong on all this - I need some experience to get a better idea of how it all works. For those of you that do have more experience I would ask this: In the past 5 years how much new & exciting stuff has cisco come out with, and how much have they dropped the prices on the old stuff. Are they coming up with new stuff that you really want in your network, or do they depend on selling retreads of old products?

Humm... It sounds like I will have to learn a lot more about the cisco stuff before being able to say much more here. It is clear that I only know a small amount of what is going on out there.... Well, if I manage to get a good real job I should be able to learn a fair bit.

It also sounds like the CCNA self study books leave out quite a bit, and are sometimes misleading. My guess is that they are intended only for beginners, and that they gloss over some of the complex issues because of this. I tend to feel that this is not a good thing, but it seems common. My better instructors in college would point it out when there were exceptions to a rule, even if we didn't have to worry about them for that class. That way, if we actually used the stuff later the exception would not come back to byte us (unless we decide to program in perl).

Some examples:

Note: the book that I am looking at is the 2005 edu. of the CCNA self study guide, ISBN 1-58720-083-X

-The prior mention of the book saying that you can't edit numbered ACLs (page 465)

-ISDN could well be viable for voice; don't know what is going on in the telco area. However, chapter 10 acts like I can expect to be using an ISDN BRI as a serious backup network interface on a router.

-There was never any serious discussion of other ways of configuring a router. I agree that using a command line can be much better way of doing things, and the CCNA should mostly focus on it. However, they should at least cover the basics of other ways to do it. I can just see someone who just got a CCNA saying in an interview that they have no clue about the http server or windows configuration programs. This could really look bad for both them and cisco.

Thanks for the various comments! They give me some idea of what to keep an eye out for, and give me a better idea of problems to look for.

One other question. I am thinking that the CCNA self study books will not make good reference books, and that when working with cisco stuff I will want a real reference book instead of a beginners instructional text. Should I just ebay these books and get some other cisco book when the time comes, or will they be useful in the future?

So, I need to do some plotting. Ideally on my mac at home (G4, OS 10.4). In this case, a pile of (X, Y) coordinates on a semi-log graph. And every occasion that I have to do this I run into the same problem: it seems that just about every program that I have to do it stinks.

When I attempted using a normal spreadsheet program (NeoOffice 2 and appleworks 5) they would refuse to take the data as (X,Y) points. Instead it would assume that each column is a data point with respect to the row number. I attempted using M$ excel 2007 (in a computer lab - thankfully I do not have my own copy), but it was highly un-cooperative. Not only did it have the above problem, but also it would do log on the Y axis, but not the X axis. Maybe it could be beat into submission, but I was not able to figure out how. It did not help that when I did a web search most of the results were for older versions of excel. Grrr...

Sometimes I use CricketGraph, which is a very old Mac program (late 80s) that does a good job for a lot of stuff. It even runs under classic on my G4! It does a great job for basic (X, Y) plots. But it does not do any log sort of plots.

Anyway, it seems like plotting is such a basic, easy, task that there should be lots of good applications out there to do it. If one exists, I would rather find and learn a good program. Hopefully one that is free/opensource, or at least does not cost much. Any suggestions?

Ah, the classic Maple vs Mathmatica vs the not yet mentioned Matlab debate. I do indeed have access to these via a computer lab. I keep thinking that I should learn how to use them, but have not yet felt like taking the time to do so. Only if my long ago Calc 1 lab taught us that instead of making us solve worthless problems on graphing calculators, and then do group presentations on them. Anyway, maybe they are the best route. Though for what I am doing they feel a bit like sand blasting a soup cracker.

Anyway, the good news is that it turns out that CricketGraph actually let you do a log graph! I just had not figured out how... Funny how a little Mac graph widget written in 1988 works so much better than the latest M$ excell.

So, I guess I will get a tad more out of CricketGraph. But I still need to keep my eyes open for something newer.

Oh, and I do need to give GNU plot and R-lang a poke.

It does a very good job of graphing intuitively. Probably because it is a good example of Keep It Simple, Stupid.

Sad to hear that they got eaten. Thought I had figured that something terrible had happened to them. I think that the best I have managed was v 1.3 from Macintosh Garden.

What's wrong? Being a tad younger than said items, I don't have any to dig out, thats what.

-----
Ah, Grapher. Maybe I should fiddle with it some more. I took a brief look, but was unable to get it to do semi-log (just log & log-log). And its point-entering interface is a pain. But if it will take a CSV it could be usable.

I also appreciate having this site around. It makes a good place to ask computer questions and discuss old hardware.

Admittedly, I mostly use it as a less profitable but far more pleasant place than ebay. Though, now that I have a real job, I am a lot less interested in reselling old hardware and more interested in keeping my excess junk to a minimum.


A little sad to see the O350s go, but quite understandable.

At work I am doing a fair bit of Java GUI programming. To learn Java I have been using a mix of the web and the 2 Core Java books. I have started making progress on finding ways to improve my GUIs. I started looking at GUI Bloopers 2.0 for some good user interface tips, and I am thinking about getting Rocket Surgery Made Easy for some more usability. There seems to be more material that I can ever read on designing good, user friendly, GUIs.

But doing it in Java is an issue. I often end up thinking that there should be a better way to do something, but I can't figure out how. And I hate saying "yea, this interface bug is stupid and it shouldn't be that way, but Java made me do it." Granted, my project chose Java as the language that we are using, so the Java-induced issues are more their fault than mine. But it <i>is</i> my job to do the best I can within the constraints of Java. So, what are the best references that you know of for writing GUIs in Java? Good web sources are fine, but I often find that a well written book is far better than a web source. And if a book actually improves my work, or makes it happen faster, it is well worth the cost. But I don't want to waste money on a useless book.

Well, it is and is not their fault. Java deserves some of the blame. But, depending on the exact issue, it could be something that the developers can and should have fixed.

An example:

My graph window has a button to export the graphs to a .png file. So, when you press it, an standard file selector pops up. You select a directory, enter a file, or select one for overwrite. Fairly standard. Under the hood, I am using the JFileChooser class - a standard java library/class intended for this sort of thing.

The problem: the file name really should end with a .png. So, if the user does not stick .png on the end, we should. But JFileChooser does not provide for this. It has a filter option that lets me limit any file they select to .png, which would be neat if I was opening a file. But it does not have any sort of filter for new files that they decide to save.

Now, anyone looking at this might suggest that I simply take the file name and run it through a simple string processor. No luck- the JFileChooser provides me with a file, creating the file if needed. It does not allow me to edit the name of the file.

There are undoubtedly ways I can deal with this. But it will take time and effort. And I have to decide between working on this, or working on other, much-needed features. So, in the short term, I am working on stuff more urgent that a user interface annoyance. I really will try to come back and clean this up. But we all know what really happens when a programmer says that...

So, it seems to me that Java interface libraries are lacking in the finer points of usability. This often has to do with fine-tuning your GUI to work in a really user friendly way. Working around / fixing them is a pain. So they end up in the final product.

Having said this, there are times when Java has a good and useful way to solve a problem, and I just have to find it.

If one exists, I would like a book that tells me how to make my GUIs better. If not, I will do my best without...