Slashdot Mirror

1. Get this book: Invaluable. Read it, from start to finish. It's that good.
2. Get this other book: also very good.
3. Check out your local Linux/BSD/UNIX user group: google is your friend for this. For instance, NYCBUG is very good if you live in New York City. Also Linux International has got a lot of conference-related announcements.
4. Pick a Linux distribution, any Linux distribution really, and try to find forums and User's group in your area. Then, do the same for another distro. And another. Lather, rinse, repeat.
5. For complete newbies, Linux Questions and The Linux Documentation Project are invaluable places to start. For more advanced advice, check out Unix Guru universe, or the O'Reilly web site.
6. Finally, do check the local university and/or community college to see if they offer some sort of training
   

Get the Unix System Administration Handbook . Some IT guys even call it "the bible" :)

However, it's plenty easy to install a hardware keylogger undetectably inside a chassis.

There are easier, safer ways for public terminal security.

i, for one, plunked down a hard-earned $20 bill and am downloading it right now. i've got an extra p2-450 kicking around and will throw it on there. $20 seems like a good deal to familiarize ones's self with an official, popular unix operating system. hey, the The Unix Administrator's Handbook (a.k.a. the purple book) covers Solaris (8, but hey...).

if you're a fan of computing and have the means, this is certainly a good deal. hell, if Microsoft put up a non-commercial version of XP for $20 i'd get that too. one can never have too many operating systems to play with.

..and if you don't think it's a good deal, don't pay for it. i love linux, but i think linux has spoiled some folks. not that that's a bad thing ;)

When it comes to technical books, I look for a good publisher first, and a good author second.

A good publisher can help a so-so author (by, for example, good editing, technical and otherwise). A really good publisher can return a rotten manuscript as unpublishable and refuse to distribute the darned thing. The publisher can only do so much, of course, but I wouldn't lightly underestimate how much.

Any author, good or bad, gets a lot of value from the publisher. No book is perfect in the early drafts. A so-so publisher may often put out a very imperfect book.

I've had a little experience (from both sides of the keyboard) with Sams at one point in its history. I don't feel they were a publisher that added a lot of value once the author's work is done. They may have gotten some good authors who put out some good books, but not as consistently as (say) O'Reilly. (I don't know how Sams is doing these days.)

A couple of replies to earlier comments:

"... In 24 Hours" doesn't mean "start at 9 a.m. Monday, be done 9 a.m. Tuesday." It means, "put in a couple of hours a day, be done in about a couple of weeks."

Some people have complained this is an introductory book that's not very deep. Fine; it's not for gurus. It might well be a good book to read before reading the "purple book"; and that's not a bad thing.

This is *the* book for someone dumped into being a sysadmin. See http://www.admin.com/ for updates.

http://www.admin.com/ for a link

dave

Advanced Programming in the Unix Environment - a bit outdated, but still good; does not cover kernel internals
The Practice of Programming - good tips related to style, algorithms & data structures, debugging, etc.
UNIX System Administration Handbook- actually shows you how to get stuff done
Concrete Mathematics - to help you understand The Art of Computer Programming

If you have to, or want to, deal with multiple unix flavours, get the third edition of the Unix System Administration Handbook. It has yet to fail to provide me with an answer when I get stuck on a cross-platform issue.

I agree. The best advice from an experienced sys admin was to get UNIX SYSTEM Administration Handbook it has great examples of how to do the same task on BSD like unices vs ATTish distributions.

One of the most relevant problem of wireless networking, at the moment, is the way you can debug it.
I quote from 'Unix System Administration Handbook' (3rd edition):
Debugging a wireless network is also something of a black art, since a wide range of variables come into play when there are problems. In short, wireless is currently a neat toy for homes, small offices, conferences, or the beach, but it's not going to replace wired corporate backbones anytime soon.
I would like much to know about researchs in this issue of the wireless networking: if you propose a new wireless network model, it's ok, but: how can I debug it...
I have a dream: wireless networks as easy to debug as RJ45-based Ethernet ones.

Nemeth, Evi and Garth Snyder, Scott Seebass, Trent R. Hein. UNIX System Administration Handbook, Third Edition. Upper Saddle River, NJ: Prentice Hall. 2001. ISBN 0-13-020601-6.

Definitely worth picking up a copy.

Essential unix sysadmin text: http://www.admin.com/

a link would have been helpful, dont you think?

Skill / hiring test one
Skill / hiring test two

--Moo

a link would have been helpful, dont you think?

Skill / hiring test one
Skill / hiring test two

--Moo