Slashdot Mirror


Ask Slashdot: Becoming a Network Administrator?

J. L. Tympanum writes "After many years as a star programmer, I have taken a position which involves maintaining and rebuilding the in-house network of a small company. There are maybe 100 machines, a mix of blade servers running Linux and desktop PCs running Windows of all flavors. Basically, I have to learn networking from scratch. I have been given an 'unlimited' budget to buy routers, switches, etc., to set up my own little test network as part of the learning process. So the question is: what's the right strategy here? What routers or switches or other equipment should I acquire? What books should I read? Should I take classes from Cisco, Global Knowledge, my local community college, or somewhere else?"

308 of 480 comments (clear)

  1. Step 1 by nuintari · · Score: 5, Funny

    Run, run as fast as you can, and don't look back.

    --

    --Nuintari

    slashdot : where an opinion can be wrong.

    1. Re:Step 1 by RenHoek · · Score: 4, Insightful

      1) Why does the network need rebuilding?
      2) Where the hell are they getting an unlimited budget from?
      3) Why, if they have money, would they hire somebody who never did any admin work?

      I'm not saying you won't be able to do it, I'm saying you try and figure out their motives and cover your ass with asbestos!

    2. Re:Step 1 by nuintari · · Score: 4, Informative

      And then, in all seriousness.

      Deploy Juniper products where you can. Commit confirmed alone will help keep you sane.

      As for learning how this stuff all glues together and works, that really depends on how you learn. I learn by trying things, and reading the manual, not from a classroom. YMMV, but I have never seen a class that did anything short of an awful job of explaining how networking works. I rely heavily on my peers and Google for ironing out issues that I cannot solve in my lab. Consider attending talks on subjects relevant to your needs, and anything that sounds even remotely interesting. Find someone more skilled than you who can explain shit in your native tongue and attempt to osmosis some talent bit by bit. Oh, and get yourself an O'Reilly Safari subscription, a nook/kindle/whatever, and start, as my friend Jeff says, consuming massive quantities of text.

      And seriously, consider running, you are in for a long, dark road of evil.

      --

      --Nuintari

      slashdot : where an opinion can be wrong.

    3. Re:Step 1 by DWMorse · · Score: 1

      Why is this modded funny? This is insightful. There's a reason I left network engineering and went back to the medical field.

      --
      There's a spot in User Info for World of Warcraft account names? Really?
    4. Re:Step 1 by Capt.DrumkenBum · · Score: 2

      I agree, God help you.
      This takes a major adjustment in your thought process'. From now on it is not your job to do things. It is your job to make sure everyone else can do their jobs.
      Secondly get a bug tracker, or job tracking software setup as soon as possible. (I use Jira http://www.atlassian.com/ but it is rather expensive.) If the request is not in Jira (Or whatever you chose.) then you do not do the job.

      One SA to another: Good luck!

      --
      If I were God, wouldn't I protect my churches from acts of me?
    5. Re:Step 1 by MightyMartian · · Score: 3, Interesting

      What I find is that you'll start out with one plan, meticulously formulated through research and consultation and even after management has signed off on it... And then you'll find out that a half of the plan didn't make any sense or didn't in fact work the way those FAQs or sales people said it would, and the other half will be trounced by new demands from the departments you consulted because they neglected to tell you a part of their needs, or changed their minds, or read some article they read somewhere.

      --
      The world's burning. Moped Jesus spotted on I50. Details at 11.
    6. Re:Step 1 by pvera · · Score: 4, Insightful

      I don't understand why this is modded funny, it is the correct plan of action assuming the move was voluntary. If this is a programmer that is trying to bail out of a sinking ship and this was the only job available at equivalent pay, then it is a completely different issue.

      The biggest red flag is the "unlimited budget" that doesn't cover hiring a properly trained network admin, instead pushing him/her to learn the whole thing from scratch at the same pay.

      --
      Pedro
      ----
      The Insomniac Coder
    7. Re:Step 1 by rickb928 · · Score: 1

      "From now on it is not your job to do things. It is your job to make sure everyone else can do their jobs."

      Just an unwanted observation, but star programmers who don't "make sure everyone else can do their jobs." by using the systems and applications they developed are either not start programmers, or are working for idiots.

      So our OP is either already used to making sure his or her stuff makes others productive, or they are being promoted necause they have a relationship with the boss.

      Either way, I'm pretty sure this is not a promotion, despite the OP's hopes.

      --
      deleting the extra space after periods so i can stay relevant, yeah.
    8. Re:Step 1 by cdogg4ya · · Score: 1

      I'll second Juniper, if not for commit confirmed but rollback 1...they have some really nice switches these days with the EX series. This comes from someone who supports both Cisco and Juniper but the adage that "nobody was ever fired for picking Cisco" is true enough as well. I don't think you would go wrong with either.

    9. Re:Step 1 by chill · · Score: 1

      There may be, however, several who were laid off after spending that much extra money on Cisco.

      --
      Learning HOW to think is more important than learning WHAT to think.
    10. Re:Step 1 by Pvt_Ryan · · Score: 1

      Personally I would go with Cisco, we had cisco and moved to juniper perimeter & dell switchs only to now be redoing our network with cisco and tbph we are glad to see the back of juniper.

    11. Re:Step 1 by poetmatt · · Score: 4, Informative

      Underpaid, underappreciated and overworked? Get back to work!

      Network admins, unless they are basically amazing, are in for a typically rough ride through trying to get things to work, as things perceived as small changes can have enormous impact on network stability. Then you get to things like bad password policies, bad hardware policies, bad security policies, bad corporate policy and a good portion of the time network administration is just not worth the time.

      If it were $75-90k a year maybe, but otherwise definitely not worth more stress than pretty much any job that exists today including hard sales.

      Things to do: buy enterprise grade hardware, do not ever compromise on best buy/off the shelf hardware, restrict access as much as possible (and lock down ports as much as possible), make sure all devices go through a firewall (outbound) and all inbound connections go through their own separate firewall (inbound). Make sure that all requests inbound have to be requested from internal. Make sure that as much of inbound connections as possible are over a vpn if external.

      Basics: make use of forwarding, proxies, reverse forwarding, nat. Make sure that all of your DNS addresses which are assigned to computers point to internal DNS only, and that the same applies to the servers. No server should have any DHCP or DNS assignments from the local ISP.

      Redundancy: You must have it. At all levels. Check for cable backups, keep spare parts for everything - power supplies, cables, extra routers, extra server ISO's and images, extra copies of VM's, etc. Make sure you have redundant UPS's. Do not daisy chain UPS's (or maybe you can, someone else will comment- I'm no UPS genius).. Make sure things are not physically linked in a way that when one thing fails, so cascades the rest. This means UPS's with hot swappable batteries. Make sure you have multiple switches and all servers have at least 2 NICs for both load balancing and additional fallover.

      Check for shit people don't think of - check where the servers are located, what cables are running overhead, dust situation, etc. Make sure that the cooling for the server rooms is appropriate and is set up such that if the leak plate (forgot the proper term) floods it won't drip directly on the servers. Check for maintenance schedules, physical and software, check for licensing being followed, check for PCI compliance. Check security requirements for the server room, for the pcs.

      Additional redundancy: virtualize wherever possible, hardware permitting. Offsite backups, offsite hardware backups.

      Additional: prepare for hilariously large amounts of fucker trying to deal with authentication between linux and windows. Linux is well documented and complicated. Windows is well documented and complicated.

      Lastly:
      Keep at least 3-6 full bottles of hard alcohol on hand, a 2 week resignation notice, and a mini-fridge full of beer in addition, and depending on the state you're working in, maybe keep a gun on hand if you're licensed and it's legal.

      Oh and don't forget, being a network administrator has basically NOTHING to do with being a network administrator. It's more like managing a circus of crying babies who have no idea what the fuck they're doing.

    12. Re:Step 1 by dkleinsc · · Score: 2

      Also, to answer the question of what J. L. Tympanum should read, a good starting point: The alt.sysadmin.recovery FAQ.

      --
      I am officially gone from /. Long live http://www.soylentnews.com/
    13. Re:Step 1 by squidflakes · · Score: 1

      Seconded. I left networking and went in to HR, though I do miss playing on the big iron. Knowing that you can royally screw some people by intentionally advertising bad BGP is a hell of a responsibility and massive ego trip.

    14. Re:Step 1 by malraid · · Score: 2

      unlimited budget? I'll solve your problem for $120K a year.... so.... when do i start?

      --
      please excuse my apathy
    15. Re:Step 1 by lymond01 · · Score: 1

      1) Why does the network need rebuilding?

      A fair question but likely for the normal reason -- it wasn't built right in the first place. Probably grew like a fungus instead of being laid like good plumbing (with a nod to the late Senator Stevens).

      2) Where the hell are they getting an unlimited budget from?

      Unlimited probably means they have no idea how much anything costs but it can't possibly be more than $10,000.

      3) Why, if they have money, would they hire somebody who never did any admin work?

      Because then they'd have two $80,000 employees on staff when they can only afford one (see definition of "unlimited budget" above).

      My company? I'd contract it out. It would probably still cost me an arm and a leg, but my programmer would still be doing his job and I wouldn't need to pay the contract company for more than a few months and maintenance visits thereafter.

    16. Re:Step 1 by bberens · · Score: 4, Funny

      I dunno, seems like a cake job to me. As a programmer I can assure you that the problem is never the network. Just ask the network guys, they'll be sure to tell you. Never mind the trace-route, pings, and FTP client log showing 100 byte/sec transfer speed I have provided, the single green LED graphic on the monitoring tool indicates with absolute certainty that all things on the network are working swimmingly.

      --
      Check out my lame java blog at www.javachopshop.com
    17. Re:Step 1 by bensode · · Score: 1

      I think I just found my new sig ... thank you so very much!

      When considering becoming a network administrator consider this advice:

      "Lastly:
      Keep at least 3-6 full bottles of hard alcohol on hand, a 2 week resignation notice, and a mini-fridge full of beer in addition, and depending on the state you're working in, maybe keep a gun on hand if you're licensed and it's legal.

      Oh and don't forget, being a network administrator has basically NOTHING to do with being a network administrator. It's more like managing a circus of crying babies who have no idea what the fuck they're doing."

      --
      "Keep at least 3-6 full bottles of hard alcohol on hand, a 2 week resignation notice,..." - Poetmatt
    18. Re:Step 1 by h4rr4r · · Score: 2

      So you want to be calling the contractors every time a new company wants to SFTP files to you? Every time you need to allow yet another protocol out from the exec vlan?

      He needs to hire someone who knows what they are doing. Then if he still wants to learn they can teach him.

    19. Re:Step 1 by amahler · · Score: 1

      Definite second on going with Juniper. Their JUNOS is fantastic, the CLI a joy to work with and their EX switches are pretty excellent, especially for the money. My only gripe: they sound like jet engines. :)

    20. Re:Step 1 by LifesABeach · · Score: 2

      the request for help kind of sounds like someone's high school term paper that's due in two weeks...

    21. Re:Step 1 by h4rr4r · · Score: 2

      MCSE certified for 10 years now,......Everything that's supposed to be automated ends up needing watched by humans

      I think I found your problem.

    22. Re:Step 1 by DuoDreamer · · Score: 3, Interesting

      This is the best description I have read regarding the Network Admin position.

      When I started as an admin 5 years ago, the company didn't know to care about redundancy, or security. When I started, neither did I. I could build PCs, do some light programming, and had a knack for finding solutions with Google. In that time, I've replaced all network hardware and fixed the topology, expanded from 6 to 20 servers, added virtualization wherever possible, added battery backup to everything (many servers didn't have any UPS), replaced 100 windows 2000 desktops, added 100 more desktops, upgraded the domain from 2000 to 2008, Exchange upgrades twice, migrated all storage to redundant RAID on server or via NAS, maintained DAILY tape backups of all servers, network monitoring via free Linux tools, expanded the network via T1 to include 7 satellite facilities and WAPs with VPN/firewall, and locked down every damned machine so that nobody can install anything. All while providing these people and locations with 24/7 tech support and software instruction. Monitoring scripts are all Perl and PowerShell, depending on OS. All of our network hardware is either Adtran or 3Com (now HP) and I've only had one switch failure in 5 years. No training seminars or certification taken, just lots of reading.

      It pays shit, but its steady.

      FML.

    23. Re:Step 1 by Coren22 · · Score: 1

      You have a problem with him getting the only certification that matters to companies for a SA?

      If he got it 10 years ago, it was most likely the NT cert, which was a very good program and required many things I use daily. Do you know how to subnet and supernet a network? You need to know to get the MCP in TCP/IP that is required to earn your MCSE.

      --
      APK likes to ask for responses to the same things over and over. Maybe he just likes the responses?
    24. Re:Step 1 by Jeremiah+Cornelius · · Score: 1

      Also, how does one program a star? Can you use any language? Or, are you restricted to Algol?

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    25. Re:Step 1 by smash · · Score: 1

      Cant be more than $10,000? I have single switches worth 6x that. Depends very much on the numbe and type of users. 5 engineers working on 3d mining models need much more capable network hardware than 5 accountants for example.

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    26. Re:Step 1 by smash · · Score: 2

      You invalidated your advice by suggesting you arent able to distinguish between udp and tcp...

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    27. Re:Step 1 by PopeRatzo · · Score: 1

      Run, run as fast as you can, and don't look back.

      It's not really funny. I mean, it is funny, but it's not funny for the poor yob who's about to give away a decade of his life just to be able to make the minimum payment on his credit cards.

      I know anyone is lucky to have a job of any kind these days, but one hates to see someone choose this path. Hard work, long hours and absolutely no respect. Raises less than the rate of inflation, continually contracting benefits packages and then, after a decade, a pink slip so the job can be given to the next young yob who'll work for less. Shrinking budget, shrinking office space, shrinking self-respect. It's a recipe for depression and substance abuse.

      Jeez, aren't they hiring at the coal mine?

      --
      You are welcome on my lawn.
    28. Re:Step 1 by nschubach · · Score: 1

      will be trounced by new demands from the departments you consulted because they neglected to tell you a part of their needs, or changed their minds, or read some article they read somewhere.

      Are you talking about Software Development or Network Engineering?

      --
      Every time I start to have faith in humanity, I ruin it by driving to work between 7 and 8 am.
    29. Re:Step 1 by Anonymous Coward · · Score: 1

      deploy pfSense, the rest is all junk

    30. Re:Step 1 by wiedzmin · · Score: 1

      I will respectfully disagree. JUNOS is the biggest step back from Netscreen (at least in the firewall family - it takes on average 4 lines of JUNOS configuration for every Netscreen/IOS configuration line), it's very powerful - it ought to be, since they took their metro-grade OS and stuffed it into enterprise equipment, but it also is metro-bloated.

      In addition to that Juniper has not yet invented active-active multi-device multilink trunking (think Nortel SMLT, Cisco VPC, Brocade MCT) and their price makes me think that they imagine themselves to be Cisco. Their only factual benefit IMHO: they have an LCD screen :)

      Now, don't get me wrong, I am a huge fan of Juniper firewalls, even the JUNOS/SRX series, but they are honestly not even close to be mature enough in the enterprise switching space. Give them 5 more years.

      --
      Bow before me, for I am root.
    31. Re:Step 1 by h4rr4r · · Score: 2

      You missed the joke. I was joking about MS solutions requiring handholding.

      Do you know how to subnet and supernet a network?
      Yup, I am university edumacated. Even done it on real hardware.

      Any companies that want Sysadmins to have MS certs is not a place I want to work. I have a real degree and use grownup Operating Systems :)

    32. Re:Step 1 by jombeewoof · · Score: 1

      This is the best comment ever posted to the internet.

      I would like to subscribe to your newsletter

      --
      Linux Zealots: Smarter than Mac Zealots, but still zealots.
    33. Re:Step 1 by lymond01 · · Score: 1

      So you want to be calling the contractors every time a new company wants to SFTP files to you? Every time you need to allow yet another protocol out from the exec vlan?

      Well, my post was kind of tongue-in-cheek but to answer your question: no. You configure the network with your needs and specifications in mind. There aren't that many surprises: you know you need to transfer files a certain way, you know you have a webserver, you use Gmail, etc etc. Include a VPN in the specs for vendor remote access. Set up a contract that allows for a 6 month post-project free maintenance to handle bugs and forgotten instances of things.

        Cant be more than $10,000? I have single switches worth 6x that.

      When I said $10,000 that was clearly tongue in cheek. As in the PHB saying, "Do whatever it takes to get this up and running! Hear me? Whatever it takes. As long as you can do it for free."

    34. Re:Step 1 by oatworm · · Score: 2

      Ha. Ha ha. HAAAAHAHAHAHAHAAAHAHAHAH... *gaaassssp* HAAAHAHAHAHAHAHAHAHAHAHA!!!

      Seriously, most SAs (at least where I live) are lucky if they can break $50-60k unless they're working for a larger firm, and, even then, you have to work your way up to it. As I understand it, there are a few greybeard SAs making six figure salaries, but the vast majority of them are lucky if they make half that. Then again, that sort of thing happens when you can whack a tree a couple of times and have a dozen MCSE/MCITPs fall out. It's also what happens when 90%+ of the enterprise world, attracted by the promise of cheap labor and fixed operating system costs, decides to embrace Microsoft and its products with extreme prejudice.

      Why? Well, because it's so much easier (and cheaper!) to find MS-specialized help...

    35. Re:Step 1 by Vancorps · · Score: 1

      That's funny, I ripped out all the Cisco switches and replaced them with HP ProCurves and we've never looked back. Much easier to use and free software upgrades for life plus lifetime warranty. Still use Cisco for advanced routing scenarios but for switching HP has been doing well, I'm not sure if 3Com will derail them though.

    36. Re:Step 1 by elsJake · · Score: 1

      HP bought 3com ?
      We're screwed , 3com switches - never had a problem
      Hp procurve switches ? Locking up like it's crazy on the slightest electrical fluctuation , pos hardware.

    37. Re:Step 1 by mjwx · · Score: 3, Funny

      As a programmer I can assure you that the problem is never the network.

      Damn straight,

      It's never an easy job because we keep everything working so well.

      Never mind the trace-route, pings, and FTP client log showing 100 byte/sec transfer speed I have provided

      Takes end users machine, turns off torrent clients, twitter clients, RSS feeds and streaming radio on the users machine and watches the speed increase to normal levels. Finally I hit the user with a rather large wrench for wasting my time.

      First rule of net admin, The problem is always the user.

      the single green LED graphic on the monitoring tool indicates with absolute certainty that all things on the network are working swimmingly.

      Second rule of net admin: The user lies. The user always lies.

      However Nagios does not lie. Nagios does exactly what I tell it to (that includes not running torrent clients at work)

      So when it comes down to you or Nagios, Nagios wins hands down.

      --
      Calling someone a "hater" only means you can not rationally rebut their argument.
    38. Re:Step 1 by gregthebunny · · Score: 1

      Cant be more than $10,000? I have single switches worth 6x that.

      When I said $10,000 that was clearly tongue in cheek. As in the PHB saying, "Do whatever it takes to get this up and running! Hear me? Whatever it takes. As long as you can do it for free."

      You'd be surprised how often some people actually get this as a demand from their boss, myself included. It's amazing how one can build an entire infrastructure for free nowadays with open-source solutions like Zimbra*, Resara*, and Bacula*, along with a few little TurnKey appliances here or there. The only real cost is hardware which, thankfully, is getting cheaper all the time...

      * why do all of these things end with "uh"?

    39. Re:Step 1 by DarwinSurvivor · · Score: 1

      And the fact that there could possibly be a time when you would be better off using a hub over a switch...

    40. Re:Step 1 by Unequivocal · · Score: 2

      Yeah double ditto. What I'm trying to figure out is how a star programmer can't know anything about networking? Every great programmer I've ever met knew a ton about hardware and networking, all the while being deadly with software. I personally think it would be very hard to be great programmer without understanding the bricks you're standing on, in significant detail.

    41. Re:Step 1 by NFN_NLN · · Score: 1

      Yeah double ditto. What I'm trying to figure out is how a star programmer can't know anything about networking? Every great programmer I've ever met knew a ton about hardware and networking, all the while being deadly with software. I personally think it would be very hard to be great programmer without understanding the bricks you're standing on, in significant detail.

      Unless you're developing a software stack for networking why would a developer need to know about the finer details of networks?

      Generally, a software developer only needs to worry about sending packets and receiving packets. They don't need to know about VLANs, trunking, link aggregation, stacked switches, etc. That is handled by the stack.

      In fact, if you're writing software that handles link aggregation and VLANing then you probably have a SEVERE decoupling problem in the way you write software. It will probably end up being the least portable code ever written.

    42. Re:Step 1 by Orestesx · · Score: 1

      Always get a trace. Wireshark is your friend.

    43. Re:Step 1 by urbanriot · · Score: 1

      Agree with you re: HP Procurve. I've found Cisco devices to be unnecessarily expensive and no better for many tasks than lower cost alternatives. I was also previously a huge fan of 3Com but they seem to be considerably less popular these days.

      Well, that and as a reseller, I absolutely hate some Cisco small business devices that require you to register on their site and jump through hoops for required software, and the inability to create non-proprietary VPN connections... but that's a whole 'nother topic.

    44. Re:Step 1 by DigiShaman · · Score: 1

      The combo you want is SonicWALLs with HP ProCurve switches.

      --
      Life is not for the lazy.
    45. Re:Step 1 by Vectormatic · · Score: 1

      You know that, i know that, but the CEO who just had his nephew set up wifi covering *his whole house* with a $80 netgear router doesnt know that..

      --
      People, what a bunch of bastards
    46. Re:Step 1 by smash · · Score: 1

      Still. If they have 1000 users they're still going to be spending more than 10k on network hardware. Without more details, sweeping statements like "can't be more than 10k" are quite possibly false.

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    47. Re:Step 1 by Vectormatic · · Score: 1

      obviously, but that assumes some realism about network equipment prices on the executive side, which rather likely, isnt there.

      Lymond's out-of-the-blue number of $10000 might not be it, but chances are someone will be screaming "you need how much?" before that network is specced and set up how it should be..

      Also, i agree with the root post, if you are a programmer, dont get sucked into network management. You dont have the skills and training right now, and unless there is a significant pay-raise asociated with the new title, you have nothing to gain and everything to lose.

      --
      People, what a bunch of bastards
    48. Re:Step 1 by paganizer · · Score: 1

      Not necessarily. 100 machines isn't really that many, and there might not be a whole lot of complexity involved... I could see a company deciding to just have someone put on a 2nd hat.
      My General Advice: 1st heavily define the current network, find everything, map it to death. 2nd set up a test network to play and practice on. 3rd make your purchases based on what your network needs, not what would be perfect. 4th re-purposing works.
      someday, somebody who knows what they are doing will look at what you did (possibly you).

      --
      Why, yes, I AM a Pagan Libertarian.
    49. Re:Step 1 by paganizer · · Score: 1

      sounds like there is an existing network, it is possible that there might not be any big expenditures needed.

      --
      Why, yes, I AM a Pagan Libertarian.
    50. Re:Step 1 by Ecuador · · Score: 1

      How was this left un-moderated?
      Haha, Algol!

      --
      Violence is the last refuge of the incompetent. Polar Scope Align for iOS
    51. Re:Step 1 by dbIII · · Score: 1

      If you are going to do that you can use a cheaper and faster switch instead of a HP one.

    52. Re:Step 1 by dbIII · · Score: 1

      No point getting uppity. I have a real degree but instead changed fields to computing :)
      That doesn't mean I'm better at the job than you or the guy that is a pretend MS Engineer because MS thought the name would sell more certificates. After a decade the above poster is probably better at what they do than both of us who are probably doing something slightly different. An MS cert wouldn't be much good where I work, but I'd probably be crap at running whatever MS Exchange has congealed into since I last looked at it while they would know the undocumented tricks required to get it to stay running.

    53. Re:Step 1 by Unequivocal · · Score: 1

      I'm not saying that they programmed using crazy network design, I'm saying that all the really great programmers I've met (say my top 3 all-time) were very skilled in networking and in hardware. They understand bus speeds, protocols and tiered caching, and could build high performance hardware if called upon. They also understood physical signaling and packet architecture, to the point where IT network guys would often stroll over for a chat to get advice when they were designing new network layouts. They absolutely didn't *need* these skills in terms of applying them in their day to day work, but they for whatever reason had them, and I suspect it made them better programmers. Maybe my experience is anomalous.

    54. Re:Step 1 by mjwx · · Score: 1

      That is, until the Nagios box loses (partial) network connectivity and after a couple of minutes regains connectivity. A shitstorm of DOWN/UP mails will be racing towards your spool.

      That's why you make the Nagios box a dependency for all other hosts and services. You only get a single up and down alert as Nagios is considered a blocking outage.

      Been that way since version 2 at least (I didn't start using Nagios until version 2).

      Leaving flap detection enabled on a high volume public facing web site, that's a good way to get a shit storm of up/down notifications.

      --
      Calling someone a "hater" only means you can not rationally rebut their argument.
    55. Re:Step 1 by poetmatt · · Score: 1

      yeah, I thought it was a bad idea, but I wanted to confirm. I remember there was something that *was* okay to daisy chain onto a UPS then, maybe it was that you can connect a power strip to a UPS.

    56. Re:Step 1 by stewbacca · · Score: 1

      Still. If they have 1000 users they're still going to be spending more than 10k on network hardware.

      I'd hate to work somewhere that has 1,000 users and only 100 computers.

    57. Re:Step 1 by rtb61 · · Score: 1

      The most obvious answer, if your asking /. what you need to do as a network administrator then go away and let someone who knows how to do the job, do it.

      Want to learn how to be a network administrator, then get a job as a junior and learn the ropes from a skilled network administrator it will only take a few years.

      Asking for hints from the net and going the trial and error approach, will only teach the network amateur and the likely nepotistic company that hired them, the error of their ways.

      Network administrators need to be able to work with hardware, work with software and more importantly work with suppliers and work with employees. Want anybody to tell you how to do all that in one slashdot comment, then your in way over you head ie, like parent said with unlimited budget hire someone who can actually do the job and just learn how to be an effective sociopath unashamedly taking credit for some one else's work.

      --
      Chaos - everything, everywhere, everywhen
    58. Re:Step 1 by bipedalhominid · · Score: 1

      Hear Ye Hear Ye, that's my new sig too.

      --
      This aint Daytona and you aint Dale Earnhardt. So stop trying to draft on Interstate 40.
    59. Re:Step 1 by lawpoop · · Score: 1

      *Whoosh!*

      --
      Computers are useless. They can only give you answers.
      -- Pablo Picasso
    60. Re:Step 1 by bipedalhominid · · Score: 1

      Yeah man, those HPs lock up like it's closing time. Seriously though, might try a firmware refresh/update on those HPs. Just saying. :)

      --
      This aint Daytona and you aint Dale Earnhardt. So stop trying to draft on Interstate 40.
    61. Re:Step 1 by bipedalhominid · · Score: 1

      Hell yeah baby, if the light is green we are good to go. That's why they use green. Now if it's amber or red there might be a problem.

      --
      This aint Daytona and you aint Dale Earnhardt. So stop trying to draft on Interstate 40.
    62. Re:Step 1 by Quince+alPillan · · Score: 1

      Come on, Minecraft isn't THAT network intensive...

    63. Re:Step 1 by Tyler+Durden · · Score: 1

      You don't think you'd be able to get a better-paying job with the knowledge you've acquired? You're doing too much to not be well-compensated for it.

      --
      Happy people make bad consumers.
    64. Re:Step 1 by MuValas · · Score: 1

      Because it almost never is. Rather it's the programmer that plugged in a wifi router to his hub because he "just had to have wifi for this project", except he kept the default settings on the router resulting in it thinking it was the DHCP server. Or the programmer that was writing network code, had a bug, and didn't realize he was flooding the network with UDP packets - here's a hint, when developing networked software use a f'ing isolated network! Or maybe learn little things like time-to-live.

      At least 90% of the network issues are caused by programmers or other users, and nothing to do with the infrastructure itself.

      And no, I'm not an admin, just a developer like you that's f'ed up the network from time to time. It's just I learn from my mistakes and am not a frigging prima donna with entitlement issues. *insert image of dog going "WARGLBRLARGLRLLLG" at a sprinkler here*

    65. Re:Step 1 by 1s44c · · Score: 1

      deploy pfSense, the rest is all junk

      Or just use PF and get the rules you want not what pfsense thinks you want. The last time I used pfsense I gave up because pf alone is more flexible and pf on OpenBSD has the best security going.

    66. Re:Step 1 by 1s44c · · Score: 1

      "From now on it is not your job to do things. It is your job to make sure everyone else can do their jobs."

      Just an unwanted observation, but star programmers who don't "make sure everyone else can do their jobs." by using the systems and applications they developed are either not start programmers, or are working for idiots.

      His boss just gave someone with no network management experience a network manager job. It's a fair bet he is working for idiots.

    67. Re:Step 1 by 1s44c · · Score: 1

      Hp procurve switches ? Locking up like it's crazy on the slightest electrical fluctuation , pos hardware.

      Seriously?? I've never seen that and I do get power fluctuations every now and then.

    68. Re:Step 1 by Jeremiah+Cornelius · · Score: 1

      BEGIN
      FILE F (KIND=REMOTE);
      EBCDIC ARRAY E [0:11];
      REPLACE E BY "HELLO EARTH!";
      WHILE TRUE DO
        BEGIN
        WRITE (F, *, E);
        END;
      END.

      --
      "Flyin' in just a sweet place,
      Never been known to fail..."
    69. Re:Step 1 by Estanislao+Mart�nez · · Score: 1

      I'm not saying that they programmed using crazy network design, I'm saying that all the really great programmers I've met (say my top 3 all-time) were very skilled in networking and in hardware.

      Your sampling procedure seems to be somehow biased.

    70. Re:Step 1 by Unequivocal · · Score: 1

      Don't be an idiot. From the start this was framed as a question, with anecdotes to provide context as to why I was asking the question.

    71. Re:Step 1 by Agripa · · Score: 1

      Off hand, I can think of three reasons to use a hub instead of a switch:

      1. A hub can be more reliable if you have devices or switches which do not implement flow control correctly or at all.

      2. A hub may work better in applications where it is better to delay sending a packet if it would be otherwise lost because of congestion anyway.

      3. Ethernet switches with excessive packed buffering can cause problems and this is not often (ever?) adjustable.

    72. Re:Step 1 by elsJake · · Score: 1

      Yep , dead serious.

    73. Re:Step 1 by elsJake · · Score: 1

      Out of my hands , haven't worked there in a while. Anyway , we had them set on top of an UPS just to avoid this kind of crap.

    74. Re:Step 1 by DarwinSurvivor · · Score: 1

      1. I'm genuinely curious as to what types (or brands) of devices cause such anomalies. 2. Don't use a crappy switch! 3. Don't use a crappy switch!

    75. Re:Step 1 by Agripa · · Score: 1

      1. I'm genuinely curious as to what types (or brands) of devices cause such anomalies.

      These days legacy hardware would be the largest source of problems for switched ethernet implementations except for consumer grade switches.

      2. Don't use a crappy switch! 3. Don't use a crappy switch!

      I have seen quite a few embedded switches with this problem in which case replacement is difficult.

      In theory a good switch with false collision flow control should be a good hub replacement but for some reason they do not always work out. In some control systems the added switch latency uncertainty is a problem. Hubs in this case have the advantage that the sender knows exactly when the receiver gets the packet barring data errors.

    76. Re:Step 1 by bberens · · Score: 1

      Real troubleshooting tools? Oh no sir, not at any big corporation. Those types of "hacking" tools are not allowed to be used on our network by mere mortals.

      --
      Check out my lame java blog at www.javachopshop.com
  2. Don't Do It!!! by rwv · · Score: 4, Insightful

    Administering networks is best left to wizards and warlocks.

    1. Re:Don't Do It!!! by Anonymous Coward · · Score: 1

      and the criminally insane

    2. Re:Don't Do It!!! by kybred · · Score: 1

      Administering networks is best left to wizards and warlocks.

      I guess Charlie Sheen wasn't available?

  3. Sony by Anonymous Coward · · Score: 1

    Can show you how to set up a network.

    You might want to take a supplemental course for security tho...

  4. Odd choices by Apocryphos · · Score: 2

    Why would a star programmer want to transition to network management?

    1. Re:Odd choices by MBGMorden · · Score: 1

      Just a shot in the dark, but having seen such things before: the company may have had in-house software that it replaced with a commercial product, negating their need for a programmer. If the existing programming staff has a good working relationship with management, they may give them some leeway to re-purpose them into a new position rather than let them go.

      --
      "People who think they know everything are very annoying to those of us who do."-Mark Twain
    2. Re:Odd choices by 0racle · · Score: 3, Funny

      It was a very dim star.

      --
      "I use a Mac because I'm just better than you are."
    3. Re:Odd choices by yeltski · · Score: 1

      A Mexican dwarf?

    4. Re:Odd choices by mini+me · · Score: 2

      Because it is sometimes fun to do different things? I, myself, love programming, but I wouldn't want it to be my only job. Life is too short to not have fun doing all sorts of different jobs.

    5. Re:Odd choices by FictionPimp · · Score: 3, Interesting

      My story in a nutshell.

      Hired to program. Soon after system admin leaves. Server's need patching, junior admin screws up some compiles, etc so I step in and fix the server environment. Congrats, you are not a system admin (doh). A few months later, network admin is gone as well. New network guy is hired, but sucks at his job and for some reason doesn't get fired (still can't figure that out). I need the SAN to function properly, and I need the network to function properly. Congrats again, you are a network admin.

      Now the title outside my office says "Programmer", but I haven't written any programs in at least 2 years. I've wrote a dozen scripts to make my life easier, but mostly I spend my time managing, install, patching, supporting, and planning the network and server infrastructure. Somehow I've also managed not to screw it up and have finally gotten to a point where I think I might be good at this. But I miss my compiler....

    6. Re:Odd choices by FictionPimp · · Score: 1

      I wrote that without proof reading. I apologize.

    7. Re:Odd choices by squidflakes · · Score: 2

      Never, ever, ever, EVER, ever touch a router if you can't proof-read. Seriously. If I could give you one piece if advice, it would be this. Your network config will live or die by as little as a single character or number. permit udp any host 4.2.2.2 eq 52 will ensure a quick escort to the exit.

    8. Re:Odd choices by umghhh · · Score: 1

      you mean there are selfless and well willing managers like that still circulating somewhere? Hmm this could change my perception of reality or even my system of beliefs.... Possibly this also means that soon we will have cure for cancer, aids & malaria as well as world peace and I get a raise?

    9. Re:Odd choices by oatworm · · Score: 1

      To be fair, "permit udp any host 4.2.2.2 eq 53" will ensure a quick escort to the exit in most DNS-integrated directory service environments because it can nerf communication between workstations and internal assets (file servers, DCs, etc.). Outbound DNS really should only come from your internal DNS servers, not random PCs in the network. Similarly, nothing keeps you off spam blacklists better than denying outbound 25 from all machines in your network that aren't explicitly an e-mail server.

      Come to think of it, prohibiting all outbound traffic that isn't "expected" and "normal" is just a good idea anyway.

    10. Re:Odd choices by FictionPimp · · Score: 1

      Well there is a difference between my job, which I'm payed to do, and a online forum where I'm trying to squeeze out a message before lunch or my next meeting.

    11. Re:Odd choices by uninformedLuddite · · Score: 1

      Though few and far between there really are nice people out there. A few years back I was living in rented accomodation. After about five years with no rent increases we got a letter from the landlord apologising to us that he couldn't reduce the rent to compensate us for being such good tenants. He didn't increase it either.

      --
      The new right fascists are bilingual. They speak English and Bullshit.
    12. Re:Odd choices by ajlisows · · Score: 1

      That is my philosophy as well. I like to mess with all sorts of stuff.....low end development, web design, network administration... Unfortunately that keeps me from reaching my full potential in any one field. It does tend to keep me employed though. I could probably make more money if I specialized but I do not believe I would like my job as much as I do.

  5. Why? by AdamInParadise · · Score: 2

    Why did you leave a position as a "star programmer" to move into network administration? Why restart at the bottom of the ladder?

    --
    Nobox: Only simple products.
    1. Re:Why? by billcopc · · Score: 3, Interesting

      Small businesses tend to have rapidly-changing needs and few staff. If they have less development work coming in, and a pressing need to replace a sysadmin, it's perfectly sane to ask the developer if he can switch hats, given sufficient resources and support. For the employee, it keeps him in a job. For the company, it saves them from having to hire a new guy, which is neither cheap nor enjoyable, and they'd have to train the new guy anyway, which is freakin' hard when the senior sysadmin is already long gone.

      I don't think it's such a stretch, the two roles tend to complement each other quite well. A good programmer-analyst already possesses 2/3rds of the knowledge required to be a competent sysadmin. You know the shell scripts will be a work of art :) I don't know why you think it's at the bottom of the ladder, because I see it the other way around. Programmers are a dime a dozen (see China). Good sysadmins are damn hard to find, which is why I have no shortage of contracts coming in from past employers and acquaintances. Trust is a big factor, because really, the sysadmin controls access to every resource, and thus by necessity has unlimited access to all your data and equipment. Who would you trust more, some kid walking in off the street with the price tag still hanging off his jacket, or an employee you've known for years ?

      --
      -Billco, Fnarg.com
    2. Re:Why? by Ironhandx · · Score: 2

      Easy. He was interested in it, and could do a better job with it already than whoever they hired previously.

      I went from Heavy Equipment Operator, to Network Administrator, to General Manager of a Building Supply in exactly the same way.

    3. Re:Why? by gangien · · Score: 1

      Is there any software companies out there that don't have a lineup of rockstar programmers? lol

      The amount of arrogance in software development is amusing to me.

    4. Re:Why? by sirsnork · · Score: 2

      If he has to ask how to "learn networking from scratch" he can't do a better job then the last guy, and certainly shouldn't be the one they hire to fix an existing mess.

      That sounds harsh but if you don't already know TCP/IP and the basic services on top of it, you're not the one to rebuild a network. Take over maintainence of an already running network sure, get a few years of seeing a working system and how it was setup then maybe.

      Seriously good network admins got their knowledge over decades, by making mistakes and learning from them. Some made more mistakes than others sure, but you can't just pick up a book and expect to be able to do a good job architecting and building a netowrk from scratch.

      If the network is a ground up rebuild, and although thats very very rarely required, you don't want someone who doesn't know subnetting for example and setting up the DNS and DHCP and having to have it up and running in X hours isn't a good way to learn on the job.

      --

      Normal people worry me!
    5. Re:Why? by VortexCortex · · Score: 1

      Why did you leave a position as a "star programmer" to move into network administration? Why restart at the bottom of the ladder?

      I assumed that by "Star Programmer" they meant that they were programmers at Star Programming.

      As for reasons to switch jobs -- The appearance of their home-page should tell you enough about Star Programmers to remove any doubt why one would want to jump ship even if it means starting over in a dubious network admin's position that requires frequently rebuilding the entire network -- note: unlimited budget == embezzlement opportunity (not for you, for the accountant that's adding 20 bucks to each PO line item you submit).

      On a more serious note: I'm sick and tired of these hypothetical "ask-slashdot" articles -- People do have Real World, answerable questions -- let's get to those, not this BS "If I had unlimited resources and a lab to build any kind of network and didn't know shit, where should I start /b/?^H^H^H^H Slashdot?"

    6. Re:Why? by wkcole · · Score: 1

      Why did you leave a position as a "star programmer" to move into network administration? Why restart at the bottom of the ladder?

      Speaking as a sysadmin who does some network work and was a pretty damn good programmer once upon a time...

      • Programming as a job gets boring. It is very unusual to have the right employer and projects to keep the task of writing code from becoming tedious and unchallenging after a few years. A diverse midsized network never stops presenting interesting challenges, because users will always be pushing for it to do more. Users like their software stable, but they want their networks better.
      • There are not millions of highly motivated would-be network admins on the other side of the planet who would love to make $200/week and who can manage a diverse midsized network remotely with a 10:30 timezone offset and an accent that users can't understand. For a programmer in the US, that competition is very real.
      • Programming is a dead end for many people. The only ways up take you into less of being a programmer. In systems and network administration, you can make a career of handling ever larger and more capable facilities without having to move into management or starting your own company or any of the other things that programmers end up moving into when the boredom, offshoring, or need to put kids through college pushes them out of programming.

      Of course, I'm not saying that all programmers have to become admins to remain in technical jobs, but it is a very common path and there are very few programmers who are still programming into their 40's. It is also extremely useful for people managing operational infrastructure to understand the practice of software development. One of the most limiting factors for an IT technician is excessively narrow focus, so a network admin with a programming background or a server admin with networking experience or a programmer who has done desktop supoport can bring useful insights that someone with a narrow career focus won't have.

    7. Re:Why? by bigbird · · Score: 1

      Of course, I'm not saying that all programmers have to become admins to remain in technical jobs, but it is a very common path and there are very few programmers who are still programming into their 40's.

      Not true. I've been programming professionally for 20 years (in my 40's), I work with another programmer in his 40's, and many of my friends are in their 40's still programming.

    8. Re:Why? by Ironhandx · · Score: 1

      While this is true, since he's a programmer already, I imagine he already knows most of the basics and is looking more for best practice and additional equipment recommendations etc. If he can get it, some sort of best practice handbook.

      Also: Taking a chance on someone you don't know but has some paper credentials(even experience is paper credentials, I've seen people with 5 years network admin experience whose most strenuous task at their previous job was configuring a new switch) vs taking a chance on someone you know works very well, knows something about the task you're putting him into and you know he learns fast and well..... A good manager will put the second guy into the job. Not the first guy.

      If you have good staff, Its better to move your best staff into your most critical positions than try to find some random guy to do it for you. Its this basic idea that most management completely fail on, and I've seen it take down everything from 1m revenue businesses to 100m revenue businesses.

      The problem lies in the fact that management don't know anything about what they hired the guy to do, and they HAVE to trust him. If, on the other hand, you have someone you can already trust, and works well... you re-train them or throw them straight into the fire(though the second option is less appealing to most).

      Also, experienced and GOOD network admins generally aren't looking for a job. If someone has a decades worth of network admin experience and is looking for a job, if their previous company didn't suffer major layoffs/go out of business(or at least is probably headed in that direction, this is what a solid interview is for), then there is something wrong with the person. Either personally or professionally. and either case means you can't have them running your show.

    9. Re:Why? by ADRA · · Score: 1

      I don't think I've ever met really good programmers that have had little interest in programming. In fact, I'd almost say that one's lust for good programming is what elevates most dev's into really talented individuals. Are you working in a company with real programmers? Something violates the stink test with anonymous here.

      --
      Bye!
    10. Re:Why? by Xest · · Score: 1

      Perhaps because he wasn't actually a "star programmer"?

      Let's be honest, any programmer worth their salt can do network administration without having to ask Slashdot where to start. The process of becoming a good programmer in itself involves knowing where to look to find things out, an ability one could not have if they have to ask slashdot and wait on the question to be published then rummage for a reply that actually helps through the comments.

      I think realistically what we have here is a bodge job billy, who didn't really know what the fuck he was doing with software development, and so figured he'd try his hand at networking, something he also clearly knows not the slightest fuck about.

      For what it's worth I personally did system/network admin and then moved into software development, if you have to ask for help to move in the other direction you are not a "star" programmer.

  6. Step #1 by Anonymous Coward · · Score: 5, Insightful

    Hire a professional :)

    1. Re:Step #1 by scgops · · Score: 1

      Actually, rent one. Two, preferably.

      First, find a local VAR who knows everything there is to know about Juniper Networks switches, routers, firewalls, VPNs, etc. Juniper's gear is rock solid. Definitely not cheap, but solid.

      Second, find a competing VAR who knows everything about a competing brand. The obvious choice for most people is Cisco, but they will overcharge you up front on hardware and every year on support contracts. For a small business, I would instead look at HP ProCurve or SonicWALL.

      Have the resellers figure out what might be wrong with your existing network and recommend upgrade paths. Assuming they actually know what they're talking about, buy the gear from them and have them help install and troubleshoot.

      Do not try to do it all solo without professional assistance.

  7. Welcome to management by characterZer0 · · Score: 5, Funny

    1) Use your unlimited budget to hire a network administrator.
    2) Go golfing.

    --
    Go green: turn off your refrigerator.
    1. Re:Welcome to management by elrous0 · · Score: 2

      Screw golfing, with that budget he could buy a Ferrari.

      --
      SJW: Someone who has run out of real oppression, and has to fake it.
    2. Re:Welcome to management by warGod3 · · Score: 1

      3) Take credit for success, blame the admin for failure. 4) Profit!

      --
      "Be polite, be professional, but have a plan to kill everybody you meet." General James Mattis
  8. This isn't a boon. It's a curse. by Anonymous Coward · · Score: 5, Funny

    I have this job now and my girlfriend tells me I wake up almost nightly screaming. I can't help but think they're connected.

    1. Re:This isn't a boon. It's a curse. by PrimalChrome · · Score: 5, Funny

      Haven't you seen Inception? You're still sleeping....the girlfriend should have given it away.

    2. Re:This isn't a boon. It's a curse. by dkleinsc · · Score: 4, Funny

      That's clearly crazy talk. Admins don't have time for girlfriends.

      --
      I am officially gone from /. Long live http://www.soylentnews.com/
    3. Re:This isn't a boon. It's a curse. by DuoDreamer · · Score: 1

      BS! I turned my g/f into the SQL programmer so I wouldn't have to deal with that server! Now I just need to find more g/fs for the other functions...

    4. Re:This isn't a boon. It's a curse. by Lightjumper · · Score: 1

      Everyone knows Network admins cant have a life or a girlfriend..

  9. CCNA by Anonymous Coward · · Score: 1

    Read the CCNA courseware. You'll learn alot.

    1. Re:CCNA by chefwear · · Score: 1

      I second this. My local community college had a CCNA course (really 4 courses fit into two semesters) with a pretty awesome, knowledgeable instructor. I understand that not all teachers are good but the online coursework that the Cisco Network Academy is pretty thorough and Packet Tracer ( http://www.cisco.com/web/learning/netacad/course_catalog/PacketTracer.html ) helps tremendously (IOU, GNS3 aside). I would gobble that up if your local CC offers it. As far as equipment goes, Cisco is all I know and from hearing my network admin bitch about our HP switch infrastructure... standardize with Cisco. I can't speak for Juniper products.

    2. Re:CCNA by luder · · Score: 1
  10. great opportunity by alphatel · · Score: 1

    Replace everything immediately, blame upper management, and start looking for a new programming job.

    --
    When the foot seeks the place of the head, the line is crossed. Know your place. Keep your place. Be a shoe.
  11. CCNA by Anonymous Coward · · Score: 1

    Just take a CCNA course, they will have all the equipment you need to bone up on the basics. If this is only a few servers and 100 PC's, save the company some money and don't get too fancy. 3750 Switch with a few access layer switches and you are good to go. Or two 3750's running HSRP.

  12. you just need to learn one thing by roman_mir · · Score: 5, Funny

    All you need is the cloud.

    What you do is get a cloud. Just connect all your machines and networks and cables to the cloud and you will be aaaaalright.

    1. Re:you just need to learn one thing by PitViper401 · · Score: 1

      You can't forget synergy!

    2. Re:you just need to learn one thing by oatworm · · Score: 1

      Well of course not! How are you supposed to shift paradigms without synergy?

      One of these days, someone is going to develop a continuously variable transmission that provides the proper amount of paradigm at all times under all load conditions - once that happens, man, watch out! Until then, I guess we're just going to have to settle for manual paradigm shifting or torque converter-driven automatic paradigm shifting, with all the trouble that entails.

    3. Re:you just need to learn one thing by ArundelCastle · · Score: 1

      What you do is get a cloud. Just connect all your machines and networks and cables to the cloud and you will be aaaaalright.

      Dat sounds like you been 'joying da cloud a bit too much, brah. Maybe turn on a fan or what now.
      One love mon.

  13. Well well by Anonymous Coward · · Score: 1

    After many years as a star programmer,...

    Troll.

    1. Re:Well well by amanicdroid · · Score: 1

      Makes him feel better than his actual job title:

      code monkey.

  14. Why? by yeltski · · Score: 1

    Why did they ask you to do it, instead of that guys geeky cousin?

  15. Oblig. Farnsworth (paraphrased) by Jon.Laslow · · Score: 2

    Getting the core switch out was the easy part. The hard part was getting the core switch out! Hehehehehe...

    1. Re:Oblig. Farnsworth (paraphrased) by elrous0 · · Score: 2

      Here, let me show you the assorted lengths of wire we used...

      --
      SJW: Someone who has run out of real oppression, and has to fake it.
  16. Some actual advice by Anonymous Coward · · Score: 1

    If you have an "unlimited budget" and you're in charge, hire someone under you who is already familiar with networking. Sounds like you're familiar enough with the SA aspect of the job, but just need help on the networking portion. Not only will hiring someone get you going quicker, their's no better way to learn than hands on experience.

  17. Agrees with Why? by Anonymous Coward · · Score: 1

    Why not use your star programming skills to study the following book about networking UNIX Network Programming Volume 1 by Stevens, Fenner, and Rudoff?

    You should get a programming job that involves writing the embedded code for routers and switches. Taking a job as a network administrator is a step down.

    The only viable route up in Salary is to become a Solutions Architect. This would involve perhaps working towards a CCIE and getting experience in a Software Architecture role on an IP based product. Architect salaries are higher than programmers assuming that salary is a factor in your decision.

  18. HP Procurve by Anonymous Coward · · Score: 2, Insightful

    Seriously. Stay away from Cisco Gear. Overpriced over complex over hyped. Look at the HP procurve line of switches. They have very good L3 L2+ switches that handle routing for small to large networks. Take the HP networking Fundamentals In Person Class. It is one week long and provides good hands on training. Their gear has a lifetime warranty and FREE Tech support during normal business hours. Did I also mention that Software Updats are FREE. No annual maintenance. Seriously look at HP Procurve. I took a job as a Net Admin 8 Years ago at a company that was an HP shop and have never looked back or ever been dissapointed by their products or support. The 2910al is a great Static Routing Gig Switch.

  19. Ignore Cisco by nbannerman · · Score: 2, Interesting

    Forget Cisco. Phone your local HP Gold Partner - get them to put you in touch with the local HP Business Team. They'll give you free courses and training, and that is the end of that. For 100 networked devices, HP kit will do the job. I don't get the obsession with Cisco - I'm running 8 networks on 10 sites that are all HP, serving nearly 10,000 students and 1200 staff, and we've never regretted bypassing Cisco altogether.

    1. Re:Ignore Cisco by dakkon1024 · · Score: 1

      I'm not going to take anything away from HP, but Cisco is still the gold standard. You’re not going to have the marketability you might get w/ a CCIE/CCNP. The man has money, let him spend it, and the obsession is that it's a standard. People can trust a CCIE cert. There isn't a HP equivalent w/ the same prestige.

    2. Re:Ignore Cisco by silanea · · Score: 2

      Your point being? If the gear survives in a campus environment it will definitely be just fine in a corporate network.

      --
      Rudolf Hess edited Mein Kampf. He was the very first grammar nazi.
    3. Re:Ignore Cisco by nbannerman · · Score: 1

      The demands of education are higher than business. Business is one station, one logon per day. Two tops if you include lunch breaks. My networks see nearly 5000 different logons PER hour. Just one of our academies has 40+ switches, 100+ Meru APs and 1000+ connected devices. Business is easy. Education - entirely different beast.

    4. Re:Ignore Cisco by Nethemas+the+Great · · Score: 1

      Yes, you don't have to suffer the abuse dished out by students who like their torrents, video, pornz, warez, etc. etc. You have suits that like their torrents, video, pornz, warez, etc. etc..

      --
      Two of my imaginary friends reproduced once ... with negative results.
    5. Re:Ignore Cisco by morcego · · Score: 1

      I second this recommendation. For small networks, HP is much better (simpler even) than Cisco. I wouldn't bother with Cisco (except routers) for anything less than 500 computers. And even at 500 I would still not be sure.

      --
      morcego
    6. Re:Ignore Cisco by nbannerman · · Score: 1

      No, fair point that - we've got databases, CMS, Exchange and all the normal gubbins floating around as well, but you are absolutely right - they're different and we should use traffic analysis to see what the load really is.

    7. Re:Ignore Cisco by The_only_matty_x · · Score: 1

      After having dealt with HP's customer service, I would tend to agree. They are stellar and they treat their folks right.

    8. Re:Ignore Cisco by Nick · · Score: 1

      No one ever got fired selecting to choose Cisco, son.

      --
      Fuck Ajit Pai
    9. Re:Ignore Cisco by wilson_c · · Score: 1

      I couldn't agree more. I have run my networks on HP ProCurve switches and routers for the past 5 years and couldn't be happier. They're reliable, support is great, and customer service couldn't be better. Lifetime warranties for considerably less than the cost of Cisco. I also think it's easier to get up to speed with the HP kit, but YMMV.

    10. Re:Ignore Cisco by wilson_c · · Score: 1

      This guy isn't asking how to get a job, he's got the job. He wants to know where to start learning how to do the job. The value of a cert to him is nil.

      Certs in general are fine, but they only get you a certain kind of job. I don't exactly hold them against job applicants, but I do look much harder at an applicant with a cert to see if they can hit the ground running. I'd rather an ignorant self-starter who can handle many different roles than a well-educated guy who has one niche of learning. I know I can get the former up to speed quickly. All I know about the latter is that they were able to pass a test, and we don't have tests at work.

    11. Re:Ignore Cisco by guruevi · · Score: 1

      Cisco is the Microsoft of networking. Only stupid managers think that Cisco certification amounts to anything (it's slightly better than MS certificates but nonetheless useless in a real environment) and once you go that way you will be indoctrinated to think that every problem is a nail (or requires a Cisco 3800 series device) and you'll be unable to see the concepts behind it or run any unit without IOS.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
    12. Re:Ignore Cisco by hood8263 · · Score: 1

      HP Works fine for our health region...

    13. Re:Ignore Cisco by Relayman · · Score: 1

      Yeah, but on your HP network can you run SNA/SDLC and token ring?

      --
      If I used a sig over again, would anyone notice?
    14. Re:Ignore Cisco by Relayman · · Score: 1

      When Cisco sold an acquaintance a 16-port switch for $2,500, I figured that the price of gold was too high.

      --
      If I used a sig over again, would anyone notice?
    15. Re:Ignore Cisco by Fnord666 · · Score: 1

      Forget Cisco.

      I'll second that, with an extra helping of forget their training too. It has been my experience that anyone can SNAFU a network, but it takes a CCNA to really FUBAR it.

      --
      'The tyrant will always find pretext for his tyranny.' - Aesop's Fables
    16. Re:Ignore Cisco by pnutjam · · Score: 1

      yeah, It seems like most of the Cisco comments are anonyomous, the HP ones are logged on accounts.

    17. Re:Ignore Cisco by Wow8agger · · Score: 1

      HP networking expert here:
      Hey look! HP routers from SMB through carrier class: http://h17007.www1.hp.com/us/en/products/routers/index.aspx
      Wow, firewalls and IPS too: http://h17007.www1.hp.com/us/en/products/network-security/index.aspx?tab=3

      They acquired a lot of neat stuff from 3com a year ago, and at this point probably have almost as wide a portfolio of networking gear as Cisco.

      -matt

    18. Re:Ignore Cisco by Wow8agger · · Score: 1

      procurve is a good place to start. if they didn't make you license features like L3 or routing protocols like OSPF or had actual QoS it would be much better. Good luck if you need to run anything better than best effort DSCP 0. Another option, buy refurb Cisco Hardware.

      They changed that last fall - all the A series stuff comes with routing, and much of it comes with MPLS, BGP, etc.

      On the E series gear (the old Procurve line) most of the routing switches now come with OSPF, VRRP, better QoS. Next code release will also include MPLS and BGP.

      -matt

  20. Whatever you do... by Bobfrankly1 · · Score: 4, Funny

    ...don't take any lessons from anyone employed by Sony.

    1. Re:Whatever you do... by TheCRAIGGERS · · Score: 1

      I work for Sony.

      ...don't take any lessons from anyone employed by Sony.

    2. Re:Whatever you do... by flimflammer · · Score: 1

      *head explodes into a fine red mist*

    3. Re:Whatever you do... by lennier · · Score: 1

      So if I take a lesson from you, I shouldn't take a lesson from you, but then that would mean that I should take a lesson from you which means... DOES NOT COMPUTE... * smoke *

      I'll go with "True".

      --
      You are not a brain: http://books.google.com/books?id=2oV61CeDx-YC
    4. Re:Whatever you do... by Starteck81 · · Score: 1

      ...don't take any lessons from anyone employed by Sony.

      Only lessons of what not to do.

      --
      "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed H
    5. Re:Whatever you do... by Proudrooster · · Score: 1

      The network admins at Sony are probably not at fault for the trespass and data theft. I think this debacle squarely falls into the lap of Sony's rock-star programmers. Anyone with a high value target like a customer database with 75 million online accounts and credit card numbers needs to make sure all that data is heavily encrypted. If the database was clear text, shame on Sony and it's programmers. This is probably just the tip of what was stolen. The intruders probably cleaned out the source code repositories as well. Sony was totally owned and the network was so reliable that no one even knew it was happening.

      Sony also needs to ask itself, why did this happen? Was this simply organized crime or was it payback for breaking their word to customers? Some very, very clever people are using the PS3 cell processor as the core of supercomputers. The data theft might just be a cover for what was actually stolen (PS3 firmware source and O/S). In any case, Sony is going to be in pain for quite sometime as the subpoenas start rolling in from state attorney generals.

      I look forward to learning the details of how this intrusion occurred. My money is on social engineering in tandem with poor quality, buggy, exploitable code.

    6. Re:Whatever you do... by fuzzywig · · Score: 1

      Who do you think just fired all their network staff and hired the OP?

  21. Re:Network+ by Niris · · Score: 1

    The Network+ is decent for getting a general idea, but it doesn't give you many practical examples or skills. The CCNA was great for learning how to actually configure a switch/router (at least the book by Odom was. Amazing books by that guy). If you aren't planning on getting certified, you can probably go through the CCNA books in about two weeks of solid reading, and the CCNP book isn't too bad, either, if you're the same way I am and just find out you like networking. Not sure what to say on the whole server side of it though, since I haven't delved that far into that side of it yet, but I'm sure others here will have a wealth of information on where to look to learn that sort.

  22. Did you hear that? by DomNF15 · · Score: 4, Informative

    It's the can of worms popping open... You don't necessarily have to "buy" physical routers, switches, etc. These days, you can simulate pretty much any network setup you want via software and see how things work out: http://www.gns3.net/ Also, asking "us" what hardware you should buy is like asking someone what kind of computer you should buy, the question is too general and the answer will depend largely on the business/security needs of the company. Tannenbaum wrote a very good book about TCP/IP networking which you may want to read: http://www.amazon.com/Computer-Networks-Andrew-S-Tannenbaum/dp/0131651838 Aside from that, you should look into the basic requirements for network administration/security and make sure you understand and know how to apply them, the topics listed here could be a good starting point: http://en.wikipedia.org/wiki/CISSP

    1. Re:Did you hear that? by theskipper · · Score: 1

      FYI, that link is for an older edition of "Computer Networks". The fifth edition was published last year. For those interested, search Amazon for ISBN 0132126958.

    2. Re:Did you hear that? by Dishwasha · · Score: 2

      I second this. GNS3 rocks, though you'll need to find the Cisco IOS firmware. Typically you can find these with some google searching for a specific IOS firmware name. You also might want to checkout PEMU for PIX emulation. Unfortunately it hogs up the CPU so you'll want to use a CPU limiter on it, but it works really well. Settings up network bridges, particularly in Windows through loopback adapters, can be a bit challenging, but I was able to take an old multi-cpu Dell 2560 with a decent amount of RAM and Windows XP, run GNS3 and PEMU on it, and connect it to a trunk port on our switch (after restricting the VLANs of course) and we were able to use it to interview potential engineers. We had them configure routers, switches, and firewalls to our specification including seting up NATs to our external VLAN to demonstrate they knew how to do simple things like map things to a public IP, etc. We even modeled and tested a redundant ATM bridge infrastructure and got the ATM configuration just right even before the physical equipment came in to our offices for an upcoming project since none of our network engineers had really done it before. To be able to lab that out we would have had to drop a ton of money just to verify we were capable of doing the project correctly and we were able to finalize the production configuration even before the hardware was drop shipped.

      It's been a while for me, but they had also gotten Dynamips (included in GNS3) running Cisco ASA code, but it was restricted to 7.X. I'm not really sure if they overcame the issues extracting the ADSM firmware with 8.4 or not. Fortunately for me I went from being a star network admin to being a star programmer and I rarely touch the stuff anymore.

  23. Don't by Sentry23 · · Score: 1

    This is not a home network.
    Peoples work and income depend on that network.
    It might look like a great job, but even when it is a mess, it is a working mess now. You won't learn everything you need to know before this thing needs to be in the air, and you run a risk of being run out of town, lynched, or something worse.

    If you have unlimited budget, name your self network architect, follow a few nice courses, and hire people who know what they are doing to do your job.

  24. Run... by dakkon1024 · · Score: 4, Informative

    I am a 12 year veteran of the field. My official title is Sr. Technical Engineer. I work for a small (15 person) consulting firm. I’m being completely straight w/ you. Start looking for a programming job. This is the end of my advice.
    If you need to fake it for a while, setup w/ a well-respected school in your area for your CCNA. If you have no budget concerns schools w/ equipment stacks and solid instruction will beat out any other option.
    But seriously, you’re making a bad career move, this isn’t meant to be funny.

    1. Re:Run... by Kagato · · Score: 2

      My perception is Network Admin jobs are undercut significantly by offshore and H1B resources. It's tough, has lots of late nights and in the end, once the hardware is in you can be replaced by someone in an office on the other side of the world and some low-wage wiring/data center techs do the local bit. There are some high end router guys who really know switching both on the traditional networking and telephony like SS7 Switching that I think can name their own price, but they are the exception, not the rule.

    2. Re:Run... by dakkon1024 · · Score: 2

      It really depends on the “type of service” For example I work mostly in NYC, mostly as a “rent a CIO.” The cost for a company to stop using my firm is usually too high to justify the savings from another firm. That is the new firm will both be unfamiliar w/ the environment and might epically suck. As we are familiar w/ the client’s environment and have everything documented we can usually continue to charge a fair price any worry little about being undercut. Now if you provide crappy service well that’s a whole different story.
      You can’t replace a router off-shore
      My H1B guy makes more money than me. (Seriously) Americans need to realize the threat of the talented hardworking workforce outside the US and become equally such.

    3. Re:Run... by ncohafmuta · · Score: 2

      I 2nd this.
      Go back to your 9-5 job. Nobody in their right mind would start in this job with hindsight. So take ours.
      It's like having a newborn that gets up at all hours of the night, and NEVER grows up.
      Forget taking a vacation and not having to worry about anything. Maintenance at 3am, check. Network goes down at 3am, check.
      You know those few people you see driving on the roads at 3am? All admins.
      On the other hand, you're usually considered pretty indispensable. But you'll never know it because you'll be wildly under-appreciated.
      If you do do it, resist the urge to grow a beard and get fat. Resist the urge to talk down to end-users, especially since they will most certainly deserve it.
      Never make a policy 'optional'. Never recommend..require. Never assume people can't possibly be that dumb. Prepare for the worst, and then a little further. Think about the possibilities you haven't thought of. Label Label Label. Inventory Inventory Inventory. Document Document Document. Always wear a grounding strap. Dust kills. Heat kills. Don't throw water on a router that's on fire. Educate users that they shouldn't plug one end of a network cable into a wall jack and the other end also into a wall jack. Oh, and don't skimp on hookers.

    4. Re:Run... by dakkon1024 · · Score: 1

      To the "Please let us know _why_ this is a bad decision comment." Ncohafmuta puts it better then I ever could :) To add. If you can say you are a "star programmer" and you want to go back to the ground floor as a "Network Admin" You seriously need to reconsider everything you have ever thought about life. It's like being the a top pop singing, and saying, man I think I would be an awesome waste management engineer. Yea, there are problems in programming. Patches that go south, deadlines not met, etc. However, the 24 hour IV that you cannot unhook that comes w/ being a "Network Admin" is to high a price to pay. It's a job for the 20-40 something people of the world. It's never a second career choice, and you need to get out, our move up. Every 50 year old killer engineer I know of has some 20 year old working for him/her so they can live a normal life. They come in, design, run, manage, and let lil Timmy come in at 2am to swap in the new firewall. You don't want to be lil Timmy as you "second career." Even more so after being a star.

    5. Re:Run... by Bacon+Bits · · Score: 1

      This has nothing to do with wages, is my guess. This has to do with understanding where your area of expertise ends.

      This is the equivalent of asking a heart surgeon to do nuclear medicine, or asking a plumber to run the electrical wires for a house. It's got nothing to do with intelligence or understanding the ideas of basic underlying systems. It's about understand that being completely untrained in this area means you will not obey best practice, you will severely screw things up, and when you get fired for screwing things up so badly you can't fix it when something goes wrong you will make the job of the next guy 100 times harder because of the poor decisions you've made. You will make mistakes that experienced professionals will not, and you will make mistakes that a business of this size should not have to endure.

      I would say that every network admin and systems admin I've met or worked with has encountered the network or system which was designed, configured, or maintained by a computer programmer. They are an absolute nightmare and a source of endless grief. Yes, the system probably works. But it's also undocumented, not secure, none of the software outside the has been main. And the general response from a programmer will be "who cares? the system works doesn't it?". Why do I have such a hard time convincing some programmers why there needs to be a difference between dev, test, and prod?

      If you're a business in need of a network admin to redesign the network, you hire a Sr Net Admin. Know your business needs.

      If you're a computer programmer in need of work, take a job as a programmer. Know your limitations.

      This is not to say there are not equally crappy admins out there. There certainly are. But being a computer programmer, software engineer, or enterprise solutions architect doesn't grant you an understanding of systems and network administration. How can I tell? Because I have to maintain your systems that break every convention of the OS security and application model, cannot be deployed without hours of repackaging, and require runtimes and libraries that are ridiculously out of date and suffer severe security vulnerabilities.

      --
      The road to tyranny has always been paved with claims of necessity.
  25. SonicWall by Metabolife · · Score: 1

    Dead simple installations, multitude of configuration options to do most everything. Still lets you get down and dirty if you need to.

    1. Re:SonicWall by Dynedain · · Score: 1

      SonicWall, the same company that shipped expensive "small business" routers that were worse than a Linksys WRT54G, and then had the nerve to charge for more than 5 internal IP addresses on top of that.

      --
      I'm out of my mind right now, but feel free to leave a message.....
    2. Re:SonicWall by Metabolife · · Score: 1

      I didn't know about their past.

      We're running a TZ210 in our small shop of less than 50 workstations and a few servers without issue. Maybe I don't have a problem because of the simple setup we're running?

      BTW, their support is pretty good from personal experience. Although the support guy did keep bugging me with questions after he helped me with a custom config...

    3. Re:SonicWall by pnutjam · · Score: 1

      pfsense, 'nuf said...

    4. Re:SonicWall by oatworm · · Score: 1

      I think they ditched the "license per IP address" model two or three years ago. I do agree that, until that happened, it was a royal pain once you hit that limit, though.

      For smaller networks, I like Fortigates, though they can be a little... finicky to configure. Unlike the "overgrown Linksys" nature of the Sonicwall, you have to really pay attention when configuring Fortigates to do much of anything useful (do I need an "IP Address" or a "Virtual IP" to get this port forward going?). I've also been less than thrilled with their IPS products. For just basic firewall/routers, though, they get the job done.

  26. before that.. by archen · · Score: 1

    Before trying all that it would be better to inventory what your network is doing right now as a starting point. Figure out what services are running, and how the current machines are configured to connect to the network. I'm assuming this wasn't all magically done and there must have been someone who did it before you. That's where I started and I learned a lot. I especially learned that our network was done horrifyingly wrong.

    As for learning, the server type doesn't matter much (BSD/Linux) but you can learn a LOT by writing your own firewall rules from scratch (use FreeBSD myself). Not saying to do that for your company, but you'd be surprised at how much you learn from documentation, howto's and experimentation for firewalls.

  27. Hire someone who knows what they're doing by Sir_Sri · · Score: 5, Insightful

    Seriously. If you're learning networking from scratch you are not prepared to be in charge of a network with 100 computers. If you screw it up, you could mess things up for days. Start at the bottom and work your way up, or hire someone who knows wtf they're doing, you could contract in someone (there are always going to be consultants who do network around). Bring one of them in, have them go over some of it with you.

    The 'go read a CCNA book' advice isn't far off. But if you're already in charge CCNA is at least one step down from where you want to be.

    I reiterate: use your money to hire someone else. Either hire them to actually do the job and become network manager, or hire a consultant in (be prepared to see this person regularly for a year or so) to come in and help you get things going. Make sure you have people on staff who actually know what they're doing, and can tell you when you're being an idiot.

    Going from programming to network administrator may as well be going to predator drone pilot. You use computers and networks, and familiarity with computer skills is great, but they are very, if not completely different skills. And while you're at it you need to learn to be a manager, because most programmers don't learn about budgets, HR practices, setting security and devices on the network policy and all that but from the sounds of it you have to decide how to spend money.

    1. Re:Hire someone who knows what they're doing by Anonymous Coward · · Score: 1

      Absolutely - Hire an expert and learn from them!

      I am 5 years into network admin, from being 30+yrs programming. It's new, engaging, a lot of fun, frustrating and rewarding -- just like programming was/is. I had the luxury of hiring a very well certified Cisco guy (not advocating their stuff, but his network knowledge was top-notch) at the beginning and I bugged the heck out him with questions to explain how things worked from simple routing through to multi-ISP configurations, configuring T1 routers, switches, VLAN's, etc.

      This network is over 150 network devices (computers come and go, I don't count them) from switches to T1/MPLS equipment through to various servers, fiber equipment, radio links, firewalls. We have 5 sites, 2 over 100 miles apart, QoS, VOIP, Teleconferencing.

      I also learnt the cabling contractors job so I can certify cables within the buildings and replace them as needed, which helps tremendously diagnosing problems.

      Now our budget is cut dramatically I have to manage all this by myself, which is quite do-able with the right monitoring tools and in the last 16 months I've not hired a single outside networking contractor.

      Don't bother with the qualifications, they will only help lock you into a single vendor and unless your network is HP Only or Cisco Only it will take you so far but teach you a lot of unusable stuff.

    2. Re:Hire someone who knows what they're doing by Sir_Sri · · Score: 1

      Depends on how important the computers are. When you are both the person using the network, and the admin (as most of us who are scientists end up being) if you break the network, you can't work, except to fix the network. If you're only a network admin, and you break the network 99 other people can't work, most scientists can be semi independent, but a lot of employees really aren't.

      It also depends on what the network does now. How many problems does it have? What is the staff turnover like of the non IT employees (if all the staff knows how everything works, even if you don't that reduces your time needed).

      I went from being in networking to being a scientist. (Well actually i went from being a scientists who did networking, to networking to being a different kind of scientist).

      So like I say. I'd break the job he's describing into 3 parts: Hardware, software and people/money.
      Learning the hardware is a 'go get a CCNA' sort of problem, and that will solve some of the software side of things. When you manage 12 computers, what did they connect to? And what did those connect to? Do you have load balancing issues? How many (physical) backups do you have, how are they organized? If you have wireless, do you know how to handle multiple access points that will have the same name to the user?

      Software is 'what services do these computers provide', and 'what networked tools do people use that you could break'. Some of it will be solved by by a CCNA but a lot of it How many layers of security (if any) do you need on various things, how do you implement it? How do you detect an intrusion, how do you stop it? How is your backup system organized in software? Do you know how to do a restore on all the different types of things you have backed up, how long does it take, what can go wrong ? What services are people using? If it's just 12 office computers with some shared folders that are mostly independent it's not so bad. Is your office scheduling tool running on them, how does it work? How about the mail server? Where are these services accessed from? Who had licences, how much do they cost, how are they maintained? What are you storing (how much of it is confidental)? How much, if any of it, has regulatory requirements? Who oversees compliance with those rules? What do you do if the only network admin gets hit by a bus?

      People is where cisco training in my experience failed miserably. Not that it was really trying. In CS I would call this 'requirements analysis'. What are you trying to do with all this software, and how much is it worth to you? I could spend a million dollars on IT for 100 computers, but you probably wouldn't get a lot of value for your money, but it depends on what you want. ultimately you're spending shareholder money, so you want to spend as little of it as possible to get the maximum return. What tools do people need to do their job effectively, and how much money are you wasting on using the wrong tools? If all your new employees only know how to use the ribbon in office but you have a 'no ribbon' policy from 5 years ago because no one wanted to learn do you now capitulate and let new employees have what they want, or do you try and train them on whatever you were doing before. On the thought of training, who does it? What needs to be taught etc.

      We give 19 year olds who are barely literate a CCNA after 16 months of college (what in the US you would call tradeskill training). It is by no means demonstrative of knowing how to actually be the one in charge, which is, on a technical level, something you can learn pretty easily in a couple of months if you have a minimum of brainpower and know something about computers (we spent 3 weeks showing the kids how to assemble a computer, the bar isn't high here for some of it). It's much easier of course if the last person to do it is there and can show you the ropes so to speak. The hard part is in finding out all of the things people rely on that are going to break if you shut down a serv

  28. CCNA by CerealBH · · Score: 1

    CCNA is definitely the way to go, you could take Net+ but its pretty much the same thing as CCNA, but not as proprietary, but seeing as how CISCO practically runs the background of the internet, its not such a problem. And even if you end up gettnig juniper products or something else, all the commands are very easy to pick up most router IOS's being unix based. CCNA helped me expand my knowledge ALOT, and I do mean ALOT. If you were going to get anything to test on, you might grab a CISCO 1841 router, there not cheap but if you have a "unlimited" amount id get one, o and me one :). If you sign up for a Cisco Netacad class, you get access to PacketTracer which is a router IOS emulator which is very useful and powerful and there are also other alternatives out there. Nowadays you can even make a Virtual Machines with a router IOS.

  29. Do it this way by mehrotra.akash · · Score: 2

    Configure static IP's on all the machines
    Take a 100 port hub or build it yourself
    connect all machines to it

    Enjoy :)

    1. Re:Do it this way by Kjella · · Score: 1

      Ye gods, no. If you're going for that variety, at least use DHCP and reserve IPs at the switch. Anyway, 100 people... if they're all in one location, I'd say three managed 48-port switches in a rack and you're done. The rest is cabling and learning to use the software. Also I'd keep a fourth switch has a hot spare, today losing the network is just as bad as losing electricity.

      --
      Live today, because you never know what tomorrow brings
    2. Re:Do it this way by pnutjam · · Score: 1

      do you work here...?

    3. Re:Do it this way by WuphonsReach · · Score: 1

      Pretty much. 100-desktops is probably about the point where you need managed switches. But that would depend on how much load you're actually putting on a gigabit network.

      Not much network traffic? No need to push multiple 1Gbps streams over the network at the same time? Then you can go with a baseline "smart" switch that allows SNMP queries (3com Baseline Switch 2948, or the basic HP, etc.). They're commercial-grade switches, not overly expensive, etc. Generally in the $15-$25 per port range (consumer grade crap tends to be under $10/port).

      But if you're pushing a lot of network traffic, or you need maximum redundancy, then you have to go with the switches that let you chain them together and do LACP across multiple switches. So that if a single switch fails, LACP automatically keeps the network running. The baseline switches won't do that and you'll have to spend more like $50-$100 per port (or more).

      So it depends on how big a downtime window you can afford. If an hour or two of downtime will not kill the company if the desktops go dark, then go with the baseline switches, keep a spare on the shelf, and keep spare open ports on the other switches. Use the more expensive switches where it matters down in the core of the network.

      --
      Wolde you bothe eate your cake, and have your cake?
  30. Adtran by Eg0Death · · Score: 1

    Look at Adtran switches. HP's Procurve support has gotten flaky.

    --
    Why is this thus? What is the reason for this thusness?
  31. Views from a New Entrant by imlepid · · Score: 4, Informative

    what's the right strategy here?

    Proceed with caution. Make sure you enjoy networking and that its challenges interest you. Networking is very different from programming and also different from desktop support.

    What routers or switches or other equipment should I acquire?

    I have extensive experience with HP Procurve equipment and I have been satisfied with their stuff. (In the network I manage we have about 120 HP switches.) They are pretty reasonable in price and have a lifetime warranty on their switches and routers (I just got a replacement for a part for something that was manufactured 10 years ago, no hassle). Cisco is good if you like features, have a large network, and enjoy spending money. I would avoid Netgear switches (unless you need a small desktop switch (e.g. GS108) to provide more ports) as I have heard bad things but I have no first-hand experience. Expect to pay around $1000-1800 for a good 48-port Gigabit switch.

    What books should I read? Should I take classes from Cisco, Global Knowledge, my local community college, or somewhere else?

    I would look to achieve a "CCNA level" knowledge. For a network of about 100 devices you won't need much more. You can do that by simply reading a book (e.g. the CCNA prep by Lammle or Cisco Press), self-study (e.g. books alone or with video) then trying to pass the test, or taking a classroom course with Cisco or GlobalKnowledge. The material covered in CCNA is useful even if you use Procurve devices (although vocab will be different, such as "vlan trunking" (Cisco) vs. "vlan tagging" (Procurve, IEEE 802.1Q))

    Background: I managed a network at a scientific research center (1000+ end user devices and a couple hundred servers). Its a mix of Cisco (core) and Procurve (edge). I have been working in networking full time for 2 years (I was in the poster's shoes not long ago) and with computers for about 5 years in a professional setting.

    1. Re:Views from a New Entrant by Bios_Hakr · · Score: 1

      I can second this.

      I'm a Sr. network engineer for a *huge* network. For 100 machines, I would probably have a Cisco 3750s for the core and, depending on the distribution of users, something like Cisco 2940s for access. If all the users are in one location, just stack the 3750s.

      3750s run about $7K each. 2940s are around $1200.

      Get a CCNA book or CCNA videos and start reading/watching.

      I haven't used any of the HP kit. I have used Juniper and Marconi for WAN stuff. As a general rule, just buy Cisco until you know you need something else.

      --
      I'd rather you do it wrong, than for me to have to do it at all.
  32. From one Network Administrator to another... by Kamiza+Ikioi · · Score: 1

    I'm a Network Administrator. With 100 computers, you have a nice small network to test already. First, you have an unlimited budget. Take advantage of that ASAP. Give it a couple years, and you'll be saying, "Budget? Don't I get a budget?"

    Get setup with some nice Cisco Gigabit. Probably just 1 or 2 routers, maybe 5 or 6 24/48 port switches. Next, get a HANDLE on your network. Either go corporate, or go free. Look into Spiceworks, Hyperic, OpenNMS, Zenoss, etc (spiceworks actually has a nice community of Network admins that you can talk to and ask questions).

    Next, get your anti-virus in order. I recommend ESET. (I assume with 100 computers, you're also the Systems Administrator, that's typical).

    Focus on security, security, security. I'd get OpenDNS on it immediately, lock out malware sites. You may find several computers already running trojans, maybe conficker. It'll be a fun learning experience.

    I'd advise avoiding Active Directory or anything Microsoft. Then again, that's laughable advise. Good LUCK avoiding AD! And then, good luck with your windows users not getting pissed because there are no policies on the Mac users! (Just remember, control them at the DNS/router level, and you'll be fine. Active directory is good for pretending like you're doing something that looks important.)

    Most importantly, go to community college. Get a degree if you need it, but at least pick up some certs. They're not worth anything, but you'll learn. Hell, if they're paying, free education is always good.

    Screw books, you don't have time for books. Go to some SANS Institute workshops (unlimited budget will cover that) and learn some hard core skills. College and workshops will give you real hands on experience no book on Cisco Routing will do.

    Now, get ready to crawl your ass over rafters and in dank dark closets. Get ready for your finger tips to bleed as you make Cat-5e cables by the hundreds.

    Get ready for the wake up calls at 4am on a Sunday because your email server is unreachable. (You got the budget, plan a cloud failover now... hell, plan everything failover now. If it's not on VMware or HyperV or Xen, make them buy all the hardware you need to get it there. Remember, it's all YOUR fault!)

    But it's worth it. Because, in the end... you are God on your network. Just remember, benevolent gods get their heads chopped off. Make them fear you a little. Take away their facebook and youtube for a week on accident... let them know you can make them cry at the push of a button. Use inexperience as a mask for your mind games and plots. "Whoops! Gee, did I do that?"

    Muhahahaha...

    Seriously... it sucks. You are always to blame for everything. Eventually, you will make believe you are God and fantasize about taking away facebook and youtube... probably while crawling through your ten thousandth spiderweb pulling another wire behind you and remembering the last Bastard Operator from Hell story you read.

    --
    I8-D
  33. Get it done, then change jobs. by Animats · · Score: 2

    "After many years as a star programmer, I have taken a position which involves maintaining and rebuilding the in-house network of a small company.

    Learn how to do it, get it done, then work hard on getting a better job. Being an administrator for a small network is a miserable job.

  34. Basics by g00head · · Score: 3, Informative
    Assuming you didn't leave out VoIP or Video Conf equipment:

    1. As above, take a CCNA course or find the materials. That will give you a good basis.

    2. Read everything you can in regards to VLANs and how they work/best practices/management by hardware OS

    3. Read everything you can about switch port management (i.e., access port vs. trunk port, again relies heavily on the chosen hardware OS)

    4. Choose your hardware: If money is no object, Cisco is reliable but more upfront and much more for yearly support. HP ProCurve is a very good economical option.

    a. Either way, use two stacked Layer 3 switches for core routing with Layer 2 switches for access layer.

    b. For Cisco products, I'd recommend a pair of stacked 3750X's, with 2960 for access layer switches.

    c. Save yourself pain later - have each access switch trunk to the core stack with an aggregated trunk, one port to each half of the core stack. (if half your core stack goes down, most of your network stays up. If one line/port of the trunk goes down, whole network stays up but speed may be affected depending upon bandwidth used)

    5. Use one VLAN for infrastructure (i.e., switches, servers, printers, appliances), use one VLAN for workstations, use one VLAN for wireless if necessary.

    a. Avoid using VTP, even if it seems like a good idea to you

    b. Do all routing between VLANs on the core stack, access switch trunks should carry all VLANs however

    c. Test the hell out of your config in a lab if you have time, lot less pressure telling them that the project is delayed by testing than telling them all work is delayed because you can't find the problem on the prod network

    d. Thank god you get a test network

    4. Once everything's built, configured, and running well - BACK ALL OF THE CONFIGS UP, and repeat whenever a config change is made.

    Good luck, and you'd really better love troubleshooting problems with very little info to go on...

    --
    "I'd make a wooshing sound, but the post was so far over your head it was inaudible..."
    1. Re:Basics by Bios_Hakr · · Score: 1

      5a. Why do you avoid VTP?

      Everything else is pretty solid. I really like the idea of connecting the access to the core stack using dual links. I'm guessing you mean via EtherChanel?

      You could go with HSRP if you *really* need uptime. But HSRP can be a beast to get working properly...

      --
      I'd rather you do it wrong, than for me to have to do it at all.
    2. Re:Basics by Wow8agger · · Score: 1

      VTP is a wonderful easy to use method of propagating vlans, and pruning them from trunks, but it has some serious security moments:

      As an example, lets say you have a switch that is connected to your network, and you detach it to do some QA/Testing on it, create and delete a bunch of VLANs (which increments the revision number), and then put it on a shelf and forget about it.

      A year later you need a switch, and connect it into the network - provided that the revision number on your test switch is still higher, that VLAN information will propagate out and overwrite the rest of your VLAN database.

      Cisco VTP best practices: http://www.cisco.com/en/US/products/hw/switches/ps700/products_white_paper09186a00801b49a4.shtml#cg1c

      -matt

    3. Re:Basics by Bios_Hakr · · Score: 1

      We use VTP extensively. I couldn't imagine trying to manage our domain without it.

      We have a policy that before we shelve a switch, we blank the config. We also have standard configs on the laptops in our equipment cage. Take a switch off of the shelf, boot into ROMMON, upload the latest IOS, then apply the standard config.

      I could see a smaller shop with no real policies or procedures in-place making a mistake like that. I've heard of it happening. And I think that's why we are so paranoid about it.

      We take an additional step of whitelist pruning vice blacklist pruning. The CCNA level stuff talks about removing VLANs from trunks. Our default is to prune all VLANs from every trunk and then we allow the VLANs that are in-use downstream.

      --
      I'd rather you do it wrong, than for me to have to do it at all.
  35. personal opinion by hologrm · · Score: 1

    Unlimited budget? Have Juniper, Cisco and HP bid on your project including installation and a managed service provider to look after it. Then take the training for which ever manufacturer you choose. I would tell you to do Juniper since you are a one man shop. I have a four man team with only two senior network engineers managing 19 sites in and out of state. The Juniper gear has proven to be the least cost of operation for us and the strange stuff we try to do. And learn one command "Commit confirmed"

  36. Re:Only 100 workstations by mjwalshe · · Score: 1

    yes Cisco use a small network of 3/4 buildings and 2000 Hosts as the case study in the CCNA academy course work.

  37. Best advice by w0mprat · · Score: 1

    What to learn: Learn networking fundamentals very well before touching anything.

    What to buy: The cheapest thing that does the job and meets the requirements. Ignore anyone in sales or any geeks with axes to grind.

    Caveat: Be very very carefull in gathering requirements.

    --
    After logging in slashdot still does not take you back to the page you were on. It's been that way for 20 years.
    1. Re:Best advice by smash · · Score: 2

      Be very careful in meeting requirements. You shitty home grade un managed gigabit switch is not suitable for pushing gigabit to office desktops for example. Leave home user grade shot at home. Not all "gigabit" switching hardware is equal. Pay attention to backplane bandwidth. If it's not listed in the product specs, there is probably a reason for that (i.e. It's shit :-P)

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
  38. Re:Only 100 workstations by ninthbit · · Score: 1

    If you want to get fancy with it, you can VLAN the workstations from the printer and the servers, then setup some ACLs. Thats the ONLY reason I say it's diffrent than a cable router.

  39. Ramp Up Slowly by boogahboogah · · Score: 1

    As an applications programmer I can tell you that we have network consultant's at our customer's sites. They get paid T&M, and little config issues we handle ourselves.

    By getting network support from a third party:
    1) Network will stay up with few hiccups
    2) Your transition to Network Dude will leave you with hair on your head
    3) Your education will not interfere (too much) with the operation of the network
    4) You'll get book learning and practical OJT
    5) PHB won't have to bitch about downtime or cover his own butt

  40. Is networking the easiest thing to learn ? by lsolano · · Score: 2

    I do not understand why people underrate networking as if becoming a network admin can be done just with a " 'unlimited' budget to buy routers, switches, etc., to set up my own little test network as part of the learning process".

    Seriously. Have you asked yourself for example, who's going to do the troubleshooting? Yourself?

    Think about a network admin that asks here what computer, software and books need to buy to become a "star programmer". What would you answer?

  41. Re:The collective wisdom hath spoken. by grub · · Score: 1


    I have supported developers for over a decade, and rarely have I found one with a grasp on how hardware, drivers, network stack and logical and physical layers work.

    I can assume they're Java developers.

    --
    Trolling is a art,
  42. Unlimited hey? by smash · · Score: 1

    Hire consultants. Buy juniper. Point finger if it breaks. If you have no experience, getting dropped in the deep end is a recipe for failure. Read up as much as you can to get an overview of the concepts involved but leave implementation details to someone who knows what they are doing. Then learn by example. Maintaining an existing well set up network will keep your hands full enough for a while.

    --
    I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
  43. This is how I read it... by canadiangoose · · Score: 5, Insightful
    Dear Slashdot,

    I'd like to become an expert in a field in which I have no experience.

    It takes many years for most of the folks working in this field to gain the knowledge required to be effective, but I am very, very smart. So much smarter than most people, infact, that it shouldn't take me more than a month or two to get a firm gasp on things.

    There's just one small problem that is preventing me from teaching myself everything that I need to know to be able to do my job well. See, I'm not smart enough to know how to even begin to teach myself anything about this field. I'm sure if someone could just point me in the right direction, I'm quite sure that I'll be able to make sense of things.

    Also, which vendors provide "easy" buttons on their gear?

    Please advise.

    MrGenius

    --
    Never eat more than you can lift -- Miss Piggy
    1. Re:This is how I read it... by Sir_Eptishous · · Score: 2

      If there were mod points for Snarky you would get em.

      --
      We play the game with the bravery of being out of range
    2. Re:This is how I read it... by daveywest · · Score: 1

      I read it a little differently:

      Dear Slashdot,

      How do magnets work?

      Sincerely,

      I own a horse

    3. Re:This is how I read it... by theskipper · · Score: 1

      Luckily there are mod points for insightful.

  44. Post setup: monitoring by mitayai · · Score: 1

    After you get it all set up, check out Zabbix as a free, open-source cross-platform option to monitor them all ( http://zabbix.com )

    1. Re:Post setup: monitoring by uninformedLuddite · · Score: 1

      Wow. Just wow. Maybe he should piggyback a bunch of d-link routers too.

      --
      The new right fascists are bilingual. They speak English and Bullshit.
  45. Re:Only 100 workstations by Nickodeimus · · Score: 1

    This response is misleading. A network with 100 workstations could easily have dozens of L2L VPNs and attendant routing, access lists, and natting involved. I should know, i work in such an environment. 115 employed end users, maybe 800 federated end users, 37 servers, 7 routers, 2 offices with 2 firewalls in each, multiple internet connections, multiple point to point connections. Honestly, the first thing you really need to do is study the network thats in place. What hardware is in place. Generally I lean toward Cisco because its solid, dependable, and has good tech support available. However Cisco is not a one size fits all pile of equipment. For instance, if you are working for an ISP then you probably want to go with Juniper. Clearly, you want to steer clear of the Small Business and SOHO lines of hardware like Linksys and Netgear. Once you have a grasp of the layout, networking connections [i.e. connections within your building (LAN), from your building to the internet (DIA), from your building to other offices, if you have them (WAN), and from your business to other business (typically VPN, but can be others) then you will have a grasp of what technologies you're looking to learn. Hitting something as broad as CCNA or other certifications doesn't get you up to speed as rapidly as possible on your environment. CCNA has a wide range of topics that you may eventually need to know by may never ever see. Focus on what your environment holds and learn those technologies. You may want to hire an outside consultant to assist you with this aspect if you have no clue as to how to go about such discovery. This should probably take less than a day for a knowledgeable networking consultant to discover, and some time more for him/her to parse. Once you have this list, including the models of the hardware you have you should begin digging into those technologies. Top of that list needs to be IP addressing and subnetting, if you don't already know those. Follow this by routing - especially if you have an environment like mine that includes multiple routers. You need to have an understanding of IP addressing and subnetting to get your systems talking to each other. With that, DHCP and DNS are are essential server services, so you should probably learn about that right from the start. THis just brushes the surface and doesn't include things like SSH, Telnet, L2L VPN, remote access VPN, and the various routing protocols. Not to mention certificate services, encryption standards, etc, etc. I found that this book has a good overview of a lot of the technologies. Its pricy, but to me it was worth it just to get a very good grasp of how everything from LAN to WAN to Wireless fits together. http://www.amazon.com/Telecommunications-Data-Communications-Handbook-Horak/dp/0470396075/ref=sr_1_1?ie=UTF8&qid=1304626364&sr=8-1 Additionally, there is a veritable mother load of data available on the web. Cisco is really good about having detailed configuration manuals for their products that are publicly available.

  46. Fun times.. by mrmud · · Score: 1

    1) Cisco is naturally the big boy. Most companies come close to emulating ios. Learning it won't hurt.

    2) I wouldn't do courses. Based on (admittedly old) experience, they are a joke.

    3) Sign up to NANOG, it won't help your knowledge directly, but will keep you updated on the latest trends/concerns/flame wars.

    4) If you thought there were "more than one way to do it" in programming, you are going to love networking.

    5) A host is a host, From coast to coast And nobody talks to a host that's close, Unless the host that isn't close Is busy, hung, or dead.

    --
    -- MrMud
  47. Been there, done that by McDutchie · · Score: 1

    I did pretty much exactly this, starting in 2004. It looks like you have the opportunity to make this fun for yourself. Show some initiative and try something new. Off-hand, my advice would be:

    1. Keep it simple, stupid. For a network that small, consumer-grade routers in combination with a few medium-grade switches will do fine.
    2. Screw the cloud; host everything yourself. You don't want confidential company data on computers managed by strangers.
    3. If non-Windows desktops are acceptable, I've had great success with Linux in combination with Linux Terminal Server Project. Saves boatloads on licensing costs and desktop hardware. You get to centralize all the management for free. LTSP comes integrated in Debian and a few other distros. There's a learning curve but it's very much worth it. XFCE makes for a good lightweight end user desktop environment.
    4. Even where Windows desktops are required, have all the network services run on Linux. No hassle with licenses and restrictions.
    5. Become fluent in Linux/UNIX shell. It's convenient and very powerful.
  48. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  49. Unlimited? by PPH · · Score: 1

    $100K for me, $100K for you and we pay the H-1B guy from India $20K to run the thing.

    --
    Have gnu, will travel.
  50. Here's what to do. by Stargoat · · Score: 5, Interesting

    I'm buried so far down here, I'm sure no one will read this. But here is what you need to do.

    1. Before you begin, attend a Cisco / Global Knowledge CCNA bootcamp. You may not leave able to program routers like a master, but you'll learn how networks work.

    2. Visit every PC, Server, Router, Switch. Put eyes on everything. Create a master spreadsheet. Document model numbers, IP addresses. Create Visio documentation of the way your network is set up. Document everything. You need a good deal of cabinets to store it all.

    3. Decide what is the most deficient part of the network, fix it with the simplest solution. If you're using hubs, buy switches. If the routers need to be rebooted constantly, buy new routers. Above all, keep it simple. If possible, stay away from V-Lans, encryption software, Linux, or anything else complicated. Do this every year.

    4. Buy one third of the total number of PCs of the network plus ten percent. Buy only one model. Create a central image with Acronis and modify that image as necessary. Deploy these models. Repeat for the next three years.

    5. Outsource security. That way, when it breaks you can blame someone. At the same time, make sure you can monitor security to prevent breakage.

    6. If possible, outsource your main application. You don't want to support the product that everyone in the institution depends on. You need to keep the network up, not software.

    7. At the end of year one, bring in a network assessment. Tell the assessor what he needs to find before he arrives. Use that the next year to justify your new purchases.

    8. Make sure you stay friends with the president / CEO. When it is necessary to reorganize the server, etc, it will be necessary to have his good will.

    9. Be prepared to work like a sunuvabitch for two years. Take your spouse / GF out when you can.

    10. Don't let them make you program again. You're a network admin. You cannot support your old programming team.

    --
    Hoist Number One and Number Six.
    1. Re:Here's what to do. by spliffington · · Score: 1

      I read it and I found it very informative. There's obviously validity in the "hire a professional" argument, but I'm not a network admin and would like to learn how they work/ good practices, which is why I'm reading this thread. Thanks for the tips!

    2. Re:Here's what to do. by ilikejam · · Score: 1

      All excellent advice except one thing: Do not, ever, use spreadsheets for network configuration data. Network config matters, and shit that matters goes in a database.
      Something Free like OpenNetAdmin is a very decent start; but if the OP has an unlimited budget, I'm sure there are plenty of capable commercial applications which will manage the network config much better than a Bunch Of Spreadsheets.

      And that is all I have to say about that.

      --
      C-x C-s C-x k
    3. Re:Here's what to do. by Bios_Hakr · · Score: 1

      I disagree about the database thing.

      1. Create a spreadsheet with your networks. It should have headings like this: IP, mask, DNS name, use, user, phone number.

      2. Separate those into subnets.

      3. Print that out and put it into a binder.

      4. Use something like Solarwinds to map your subnets.

      5. Use a *pencil* and fill all the information into the binder.

      6. Photocopy the binder and leave one copy at home, one with your boss. Carry the original with you at *all* times.

      7. Update often.

      I have used databases, spreadsheets, specialized tools like SolarWinds, Orion, HP OpenView, etc. Nothing beats a printed binder that is always with you. Using a pencil will ensure that future updates, additions, or deletions will not mean printing a brand new sheet.

      --
      I'd rather you do it wrong, than for me to have to do it at all.
    4. Re:Here's what to do. by ilikejam · · Score: 1

      Points 1,2,6,7 are exactly why you want this stuff in a DB: backed up, replicated off-site, consistent, and up-to-date.
      If you want a hard copy, extract it from the DB and print it out (and always with a datestamp/index so you know how far out of date the ex-tree version is).

      --
      C-x C-s C-x k
    5. Re:Here's what to do. by starfishsystems · · Score: 2

      The above advice is the best of the lot, in my opinion as a network architect. But you will want to ignore items 4, 5, and 6, since they are not about network design or operations. Instead see item 10. Unless you're agreeing to wear multiple hats, it's not your job to do system administration or application project management. But the rest of the advice is good.

      Network engineering can be quite satisfying, not in the creative way of software engineering but in a more deliberative, methodical way. For all that the space of possible solutions in network engineering is surprisingly large, the space of architecturally sound solutions is much more constrained. But you don't know that yet. The above steps will buy you the time you need to figure out the difference. Proceed cautiously. Know where you are before you decide how to move to somewhere else.

      I'll recommend one item which I think you'll like as a software developer. Use SNMP. Once you've got item 2 down solid and you know where everything is, you really really want to know what it's doing. That way, if it starts doing something weird, you have a hope of understanding why. So you set up Cacti or MRTG or whatever to poll the devices and maintain historical data, and you graph that data and you keep an eye open for weirdness. The basic metric, of course, is traffic through each interface. You can see how close a given link is to saturation, and under what daily conditions. There are other useful metrics that you'll get around to eventually. I can't believe how many people who call themselves network engineers don't do this. Maybe it's because they don't think programmatically. That's your advantage.

      --
      Parity: What to do when the weekend comes.
  51. what an oppurtunity! by trybywrench · · Score: 1

    Man I would be learning everything I could get my hands on/enrolled in. Having said that, don't over do it. A good network is a simple network and don't forget that everything will be running on it so if it breaks everything breaks.

    They guy above who said to contact an HP Gold Partner has the right idea but do the work yourself that way you learn it instead of just contracting it out. From the sounds of it it's not like you're going to be buying insane networking gear that supports OC-19whatever so sticking with a firm like HP and taking the best practices route is the way to go.

    --
    I came to the datacenter drunk with a fake ID, don't you want to be just like me?
  52. Re:Slashdot Consultants, LLC? by robot256 · · Score: 1

    Advice is one thing but this is a "do my job for me because I'm not qualified to do it" question.

    You mean this is not Slashdot Consultants, LLC?

    I'll believe it when I see their name on my paycheck ... *checks mailbox* ... nope, nothing yet.

  53. Take that unlimited budget, and... by Gaerek · · Score: 1

    ...hire someone who actually knows the job you've been hired for. The fact that you have to ask these questions tells me you are the wrong person for the job.

  54. Re:it's a long road and your tires are worn by billcopc · · Score: 1

    Not true at all. Programming is about logic, and logic is what you need when you're troubleshooting a tangled mess of gear and wiring. Then you write the scripts that monitor and maintain everything, that's pure programming.

    Routing can be a headfull, but again, if you've already nurtured a critical, information-heavy thought process, it should be easy to pick up.

    As someone who made the same transition over a decade ago, I can't quite remember how I did it, and it certainly had its moments, but I always had the tools and problem-solving skills to work my way through problems. Can't quite nail a concept ? Write a program that tests your assumptions. I learned more things that way than from any books or teachers.

    --
    -Billco, Fnarg.com
  55. After your CCNA and all that crap by rickb928 · · Score: 1

    Get comfortable with Wireshark. And read all the Laura Chappell you can find. She's my go to for network errors, diagnosis, and everything that goes on the wire. Just be thankful you don't have to learn Token-Ring. No one will let you alone for a moment without pointing out to you how much it sucks.

    --
    deleting the extra space after periods so i can stay relevant, yeah.
  56. Wow, you get to buy test equipment? by skids · · Score: 1

    Many are not so lucky.

    My advice is this: do not trust the vendors. Do not trust the documentation. Do not trust that there even will be documentation to mistrust.

    I don't care if you are buying the top shelf gear from the leading vendor, do not assume that gear will be competent.

    Figure out what features you need, and if you need a feature, test it, and test it thoroughly on live hardware. Test all possible scenarios you can conjure. You'll learn how to use the feature better than if you merely read the manual, and save yourself a lot of sleep.

    As far as training Cisco's is very overwrought, and is at least half sales pitch.

  57. You're an idiot! by Nethemas+the+Great · · Score: 1

    As with many, I question the "star programmer" bit but that angle not withstanding. Take you're budget and hire a consultant. Position yourself as supervisor to this group. Mean while if you really are a "star programmer" go look for another job.

    --
    Two of my imaginary friends reproduced once ... with negative results.
  58. Be careful by Firemouth · · Score: 1

    The last sony network team didn't do too good!

  59. Certification by CmdrPorno · · Score: 1

    Since you have an unlimited budget, get certification from Microsoft and Cisco. You might actually learn something, and it makes you more valuable at performance review time (valuable to your current company as well as potential new employers).

    I'm curious--and jealous--as to how the OP landed this gig with no experience.

    --
    Sent from my iPhone
  60. Serious Advice by hodet · · Score: 1

    Talk to your manager/director/ceo....whoever. Try and understand what it is they require of their network, what are your real deliverables on this project? What problems are they trying to solve. Find out what your budget really is because unlimited sounds a little vague. Get a project plan together and determine what is you actually need to do. Hire some expertise to help you with the technical aspects of the project. Throw away your books, you need to manage the project and that means hiring the right people to carry out the tasks and reach various project milestones. You will be too busy managing these people and keeping the project on time and on budget to really waste time learning all aspects of network administration. There is no shame in contracting out aspects of the infrastructure (maybe you keep maintaining servers in house and contract out switches, routers, firewalls etc etc). You sir are now a project manager, not a network administrator. Good luck.

  61. Re:Only 100 workstations by mlts · · Score: 1

    I'd recommend splitting the workstations up by division and putting them on separate VLANs. Why?

    1: The guy in HR runs a Trojan which phones home. Instead of nailing the complete company, it only nails his department.

    2: Be able to comply better with Sarbanes-Oxley. For example, I'd probably put machines in Finance on an isolated network that isn't connected to the outside world, but have some Citrix or Terminal servers for Web browsing. This way, internal stuff can't make it out even if boxes do get compromised.

    3: You want your IT guys to have access anywhere, but other departments not to be able to touch servers that are internal, such as the TSM/NetBackup/Networker box.

    Oh, and the OP better think of other things other than just basic connectivity:

    IDS: You want to see where the nasty 0-day malware infection is coming from.

    IPS: You want non-zero day malware infections stopped, or the guy who thinks he is awesomesauce and is running nmap dealt with.

    Site blocking: Depends on the corporate culture. Pr0n blockers may be needed, or just something to block the rogue sites used for phishing, or known botnet C&C machines.

    SSL interception: BlueCoat may be needed depending on what rules/regs/contracts are put in place.

    Transparent proxy to ask for a username/password before Web access is granted. This will help with keeping malware from phoning home, as well as provide audit trails (which are needed in some businesses). Other places don't need this. It is all up to the corporate culture.

    Just having connectivity is just a tiny part of the entire picture. You have to have security and the ability to have software automatically isolate machines.

  62. Re:I'm a Network Engineer by pnutjam · · Score: 1

    oops, I posted that anonymously...

  63. Cisco Networking Academy by UbuntuniX · · Score: 1

    ...is a great resource, if you only ever want to work with Cisco products.

  64. Suggestions - Have Backup! by OMA1981 · · Score: 2

    There doesn’t seem to be many serious responses here. My recommendations

    Find a consultant who 1) knows his/her stuff, 2) you trust, and 3) is willing to sit with you and explain what they are doing, why they are doing it, and review your work when you make changes (preferably before the changes are made). You don’t want to be learning from someone who is wrong and you will need someone to fall back on when things get really hairy (and they will, I promise). You also need to set expectations with the consultant early. They need to understand your intention is to manage the network yourself and not rely on them 100%. Consultants often are reluctant to release passwords to someone if they believe you will only screw stuff up and then likely blame them for the failure (though not to imply that a consultant would ever forcibly withhold passwords). They need to know up front this will be a dual-managed environment and have an accurate understanding of your expectations of them.

    I highly recommend a few structured professional training courses. Go to their training facilities for the classes (minimize distractions while learning). Wait till you are a little more experienced before trying online at home courses. For intro level network stuff, nearly any vendor will do. Pick one and roll with it. Just don’t get too caught in certifications. They are great and all but won’t give you what you need to run a network. You need *practical* knowledge. Save the certifications for when you want to leave this job.

    Consume any information you can: books, podcasts, YouTube How-To’s, etc. Just remember, trying to find time during what will soon be your very busy schedule to read books and play in your lab will be tricky. Hence the structured professional training above. CCNA books are considering to be a good starting point. Just remember, these only teach fundementals. When it comes to "how will this router react when I make this change", those answers take time and experience to learn.

    As for the lab, get a minimum of two switches, two routers, and two firewalls. They should match or be smaller versions of what you have in production wherever possible (if the lab doesn’t closely represent production, it won’t help much when you try to reproduce problems or test solutions). This will give you enough to build most typical network scenarios. Beyond that, it’s a crapshoot on lab gear. When it comes to hardware vendors, buy what fits your need. I don’t personally care if it’s Cisco, HP, Brocade, or Bob’s Networking Stuff. I would encourage keeping the total number of manufactures low in order to maintain manageability but which manufactures you choose is a business question only you and your company can answer.

    --
    The less you talk, the more people hear you say.
  65. Learn from an expert by Princeofcups · · Score: 1

    You learn server administration and networking by doing, in particular, solving problems. You learn architecture by knowing what was done wrong in the past, and not doing it that way.

    Use your unlimited budget to bring in a contractor who has a ton of experience in the field. Learn everything that you can from them while they are available. And make sure everything gets documented.

    --
    The only thing worse than a Democrat is a Republican.
  66. Re:The collective wisdom hath spoken. by pnutjam · · Score: 1

    This is true, I see all kinds of expensive recommendations on here, some reasonable ones and some crazy ones. You can can the same performance at a quarter of the price if you hire a competent Network Admin. And spending 4 times as much doesn't guarantee you a good network experience if you don't have a competent Network Admin.

    At the very least you should consider hiring someone to look over your shoulders, there are all sorts of pitfalls and mistakes you can easily blow $10k on if you are new to the game. Of course you have to hire someone who isn't looking to butter their bread on both sides, sometimes difficult in the world of IT consultants.

  67. May god have mercy on your soul. by geminidomino · · Score: 1

    As someone who, thankfully, went the other way, here is the only advice I can give you.

    "Down, not across."

  68. Basin networking is not hard by hugetoon · · Score: 1

    With such modest setup (~ 100 hosts), if You're a real programmer, then You should be able to handle it easely.

    Task 1: learn how Ethernet technology works (for example: what a collision, why frames have certain length, why there are no crossover cables for gigabit ethernet, how 802.1q and 802.1d works). Some historical knowledge won't hurt it will give you solid bases and intuition valuable for troubleshooting.

    Task 2: learn how IPv4 work, specifically stating routing. You must become proficient with netmasks (not only those containing 255's and 0's), broadcast addresses, RFC1918, NAT, TTL, and so on. At this point You do not need to bother Yourself with such things as multicast or dynamic routing.

    Task 3 : learn protocols and related tools supporting IPv4: ARP, DHCP, ICMP

    Task 4 : learn IPTables (this is an arbitrary choice of mine, but I like it).

    Those are not steps but tasks that can be performed in parallel.

    By preforming those tasks You will discover other topics of interest.

    Once You are good with all that, refrain from over complicating things (read my other post:http://slashdot.org/comments.pl?sid=1244813&cid=28091781)

  69. Network Administrator, just what does that mean?? by David_Hart · · Score: 1

    Given the environment and the job title, I'm guessing that you will be responsible for both servers and the network. I would start out by hiring a local network consultant to review your LAN/WAN requirements and suggest a packaged solution. Given the small number of devices, this should take no more than a few hours consulting time. Meanwhile, I would start taking OS classes (Microsoft, Unix, etc.) and/or perhaps a BootCamp.

    While it is cool to learn networking, if you are in charge of the servers you will need to understand how to configure, administer, back-up, and recover these systems. Since the network won't change much once configured, you are better off focussing on the server side. Once you understand the server side, you can then start playing with networking. Honestly, with the size of your network, you aren't going to get heavily into network engineering.

    If you are truly interested in the networking side vs the Server side, then you should focus on getting a CCNA/CCNP and finding a job in a bigger company. Then again, your company could be one of the few that is poised for explosive growth.

    Good Luck,

    David

  70. don't worry about all that expensive stuff by alta · · Score: 1

    Bestbuy has ALL the hardware your need. Buy linksys or dlink or netgear. Oh, and their off the shelf computers are great too. remember lots of harddrives, that's important for a good fast server... So get some of those external cases and a bunch of USB hubs. Spindles = Good, for some reason.

    --
    Do not meddle in the affairs of sysadmins, for they are subtle, and quick to anger.
  71. Hire someone who knows what they are doing by h4rr4r · · Score: 1

    You can learn to do this, but it will take years. You need someone who knows what they are doing now. Go hire someone and have them teach you.

  72. Well.... by ZonkerWilliam · · Score: 1

    Personally I would go with Cisco, I've found them to be more reliable and easier to get the info you need troubleshooting wise. Juniper is Ok, but depending on the equipment, not as reliable as Cisco. Course wise there are plenty from tech colleges to dedicated courses from VARs, all though they tend to be somewhat more costly.You haven't explained what kind of network, ie. is this a flat topography? typically all vlans, layer 3 routing going on? it's sounds like a rather small network. If it's pretty flat then focus on the switching mostly for learning, routing is another level and could cause you a lot of headaches, both in design and troubleshooting. Then you have to throw in network security, firewalls,IPS/IDS, no matter how small a network, you will need these. It's a lot for someone to learn right off the bat, but should be doable after a few years.

  73. Re:your over your head if you are asking slashdot. by h4rr4r · · Score: 1

    Pick one desktop vendor
    Pick one server vendor
    Pick one storage vendor
    For the desktops get a remote administration client and an enterprice antivirus/antispyware client on them if you want to get any other work done.
    Lastly pick an authenticator, either MS AD, or LDAP backend and be consistent.

    All that stuff is a Sysadmins job, not a NetAdmins job. No they are not the same thing.

  74. Some resources for learning by TheBrez · · Score: 1

    Document everything you can.
    Backup configs, make sure you save them frequently when things are working.
    Get a good network management/monitoring package which uses SNMP to monitor the equipment.
    Take as many classes and training sessions as you can.
    Purchase vendor support for equipment. Cisco TAC is invaluable when the excrement hits the oscillating device. When the network is down, and the boss comes into the server room to ask when it's back up, it's much more comforting to hear that the vendor is helping you investigate the issue than to hear you have no idea what the problem is or when it might be fixed.
    Build a lab to test/learn new protocols/ways of doing things. Have a couple servers in there, as well as the same type or smaller versions within the same family. If you're running Cisco 3945 routers in production, a lab with 1720s running 10 year old code doesn't help you troubleshoot production issues or test code upgrades.
    A good podcast which covers CCNA/CCNP level topics with examples:
    http://www.ciscohandsontraining.com/
    How to backup your devices:
    http://www.shrubbery.net/rancid/
    Netdisco, good tool for network discovery and host tracking
    http://www.netdisco.org/
    Join and read network mailing lists. NANOG, Cisco-NSP, Juniper-NSP are a good place to start. http://puck.nether.net/mailman/listinfo/ to subscribe to several of those.
    Beyond that, good luck. Speaking as someone who has been doing systems/network administration for close to 15 years, you will learn something new every day. If you don't, you're not trying hard enough.

  75. A list! by GreggBz · · Score: 1

    This sounds like a tall order. I'd be scared. Buying equipment is not going to fix anything. You've got to learn the existing network before you can make educated purchases. From the scope of the network you describe, here are the basic things I think you'll need to learn about.

    Learn about routing. Subnets, CIDR, the differenec between a subnet mask and a wildcard, the difference between static routing protocols and dynamic routing protocols. Default routes. Policy based routing. Observe and document the different subnets you see in your network, figure out their purpose. Look at the default gateway of the clients and the servers. Figure out what device that represents. If you have only one subnet, your network is probabbly to flat. I'm guessing you have at least 2 or 3. Make a diagram.

    Learn about VLANS. Tagged VLANS (802.1q), Cisco VLAN discovery (if applicable). I prefer Brocade equipment for switching / layer 2. But I digress. What VLANS are in each switch and how do the physical wires correlate? What subnets run on what VLANS? If you have fiber, you have another heap of things to learn about. Learn how to make an ethernet cable.

    Learn about firewalls. iptables (if Linux), ASA / PIX if Cisco etc.. Learn the difference between access-lists and statefull firewalls. Learn how to add rules to whatever firewall you use. What networks route where and what firewalls are between the networks?

    What are the single points of failure? Learn to deal with those single points of failure. What are the entry points? What software is everything running? What are the link speeds, where does traffic go, aggregate and split up?

    Gather all the contract information for your equipment. Make a printed list of numbers for who to call about what. Seek consultation to fill any uncovered gaps.

    Look into graphing software with auto discover. PRTG is wonderful and not that expensive.

    In my experience, things don't usually break. When they do it's because:

    A.) Someone touched something.
    B.) The power went out.
    C.) Someone touched something they were not supposed to.
    D. ) You ran out of capacity (in a hard drive, on a link.)
    E.) A server got overwhelmed.

    Lastly, make sure everyone does their Windows updates :-)

    1.) If it's not broke, don't fix it. Why does this network need "rebuilt?" What's not working?
    2.) Make sure you can put it back exactally how you found it before trying anything.
    3.) Never, ever, make a change at the end of the day, or on a Friday. Come in early, real early, for big stuff.
    4.) Listen to your users. If they say somethings different, it probabbly is. Take everything seriously.

  76. Re:Step 1 /jk... ? by Thud457 · · Score: 1

    First, write two letters...

    --

    the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff

  77. The networking is only part of the battle. by spacepimp · · Score: 1

    There are a lot of variables so some answers may or may not be relevant, but take what you can use and discard the rest. I had exactly the same task set before me excepting the lack of network knowledge. I did have a lack of knowledge of their network however. 1. Look at the current points of failure. Often if you inherit something in dire need of an overhaul if you touch change or modify or fix one thing, you will break something else. It could be roaming profiles in the Novell network failing due to bandwidth bottlenecks. It could be the m0n0wall BSD that was incorrectly implemented and allowing SQL queries from the outside. It could be the fact that they are running Small business servers licenses and already have over 100 machines and employees. These are some painful memories.... 2. Take a serious look at licensing. Do you have Open Value? Are you running pirated versions of Windows 2000 and Adobe Acrobat 6 standard on every machine? Make certain you factor the costs of actual compliance in when you make a ledger of what the complete costs are going to be during the overhaul. If the company has never spent money on licensing in the past expect a shock and horror response when you tell them about how the BSA works. 3. What are the possible upgrade paths with current servers? Are they running outdated GroupWise, and they want Exchange? Good luck migrating to Outlook 2010. Are the backups working? Are they backing up the relevant pieces? Are they backing up to tape? to the cloud!? To a desktop in the corner of the Accounting cubicle? 4. If you decide to implement new servers: What is the current state of AD? 2000 2003 native? 2008 etc? You might have trouble upgrading the AD level depending on the desktop OS versions. You may need to have a contingency plan for mass desktop upgrades. If you are doing a complete ground up overhaul, be aware that you may be a good candidate for VMWare servers, and Citrix VDI/desktop solutions. Make a list of the Apps you use that will not work over a virtual desktop. Auto CAD, Photoshop, AVID. Any lag in the system and they will hate you. Don't screw it up. Make certain the SAN you buy in that scenario has enough i/o to work effectively. 5. As for the core network upgrade. Rerun the entire network with Cat 6 dual runs. Double them up so there is a lot of room to move, and be prepared for growth. Gig switches absolutely. They don't need to be Cisco, Juniper is great too. Perimeter security: leave that to an expert. outsource to a reputable a responsive company. You are not a cisco expert. nor will you become one over night. You can learn how to support them and do minor configurations but CYA. If you are investing in all new telephones, VOIP and POE be careful Cisco to be supported will need an entire core of all Cisco. (Hidden tax) Don't go go low end don't go high end, don't get a system that has a million features you won't ever use. 6. I'd suggest not reinventing the wheel. Get a powerful mail server with lots of room for growth. Make certain you cover your ass with a solid and effective back up system. Standardize hardware. 25% a year get a refresh . Full cycles every 4 years. Look at inboxes and email usage. Employees want to use it as a filing system now. Be careful who you piss off. In my scenario, there were no email inbox cap sizes or message size limitations. Try to transfer a 8GB GroupWise inbox into Exchange or anywhere, you'll see how corrupt the files really are. You might learn people get really upset when you tell them they can't run as local admins any more, Watch their indignation when you explain to them they can't install pirated software or the version of Photoshop their cousins friend gave them. They will be mad at you when you explain to them that just because you bought a new mail server that they can no longer email 600mb videos of them skydiving to the their friends in the company. When they hear they are getting a complete system overhaul they think it will mean that they can send bigger files, store their itunes collection

  78. IT Cartel... by geekmux · · Score: 1

    Small company and you have an "unlimited budget"? What the hell man, did you start working for a drug kingpin or something?

    Might want to take a closer look at the books before staying there...

    No, no, not those books...I mean the other set of books.

  79. Good luck. by rAiNsT0rm · · Score: 1

    I've been a Network Admin/specialist for over 8 years and now at a Sr. level with a large global company. Honestly, you are not ready or equipped no matter what you buy hardware or book-wise. The reason is that Networking relates to almost nothing except if you had previous telephony experience. There is a lot to learn to even crawl, WAY before test labs and equipment. My suggestion would be to start with Network+/A+ to get some background in general hardware and networking. Then do your CCNA and probably starting with the easy track. At that point you will have touched some gear and have some idea of what you are doing, that is also a solid 6 month or so commitment alone. IMO, though, I would probably steer clear of Cisco for the actual products and possibly go with HP (cheaper and life support) switches. I have used Alcatel, Juniper, Cisco, and even Dell depending on the needs and fit.

    Here is the next part of the issues though... security, addressing, VLANs, port bonding/aggregation for VMs, and even cursory knowledge of troubleshooting, punchdown tools, cabling standards, local code, etc. You WILL get most of this horribly wrong. You will have massive amounts of downtime, you will have angry users, you will have catastrophic cockups... all of these things can become very complex very fast.

    Honestly, my suggestion would be to get more into the systems side if you have the interest and slowly learn networking. If you believe you truly love networking, you will be certain if/when you pass net+ and CCNA... or you will realize you have no interest or passion for it and have some good knowledge under your belt to work on the systems side.

    --
    http://teasphere.wordpress.com - A little spot of tea
  80. Isolating devs is another reason for VLANs by dstyle5 · · Score: 1

    One place I used to work at had everyone on the same subnet. The QA team started seeing strange things with their QA systems. It turned out the devs were trying something new on the server side of things and were spamming the network with garbage traffic, causing chaos for the QA team. A few weeks later the different depts were all on different subnets.

    1. Re:Isolating devs is another reason for VLANs by mlts · · Score: 1

      Exact thing happened with a small company I worked for. The devs were testing some code to make packets for a distributed application go between machines effectively, and created a pretty snazzy broadcast storm, taking down the entire internal network.

      Thus began having multiple subnets, with development and QA having their own "padded cell" VLANs.

    2. Re:Isolating devs is another reason for VLANs by dstyle5 · · Score: 1

      The company was eventually acquired by a bigger competitor and hence their IT guys took over. The IT guy from our company said the their head IT guy "didn't like VLANs" or something to that effect. I'm sure he'll "learn" to like them soon enough...

    3. Re:Isolating devs is another reason for VLANs by dbIII · · Score: 1

      An air gap is better. Every new developer seems to have to learn about DHCP by kicking the CEO's laptop off the network unless you have something to stop them doing it. I have to admit that it is a good way to get the budget for decent switches if you can keep your job by removing the developers gear from the network before annoyance turns to rage. No time to use wireshark - just start running, find the new developer and pull their cable out.
      Give them a sandboxed test network with an air gap and they will still do it on the real network until they grow up a bit.

  81. Learning how to network by Momboleum · · Score: 1

    I found CCNA classes extremely helpful in learning how to manage networks of all sizes. The best part is that you have a budget to buy the stuff you need to practice and the opportunity to play around with the stuff you learn in class. I'm not sure where you're located but this link might help get you started. Good luck! http://www.cisco.com/web/learning/netacad/course_catalog/index.html

  82. The business... by munky99999 · · Score: 1

    Now they have a skilled programmer and net admin. Except pay probably hasnt doubled but your workload most likely has doubled.

  83. BNC by Viewsonic · · Score: 1

    Buy lots of coaxial cable and a bunch of BNC connectors.

    Anyone that trips on the cable and knocks the connector off their PC, fire them.

  84. You have an UNLIMITED budget? by shadowrat · · Score: 1

    build skynet.

  85. Novell Netware by droidsURlooking4 · · Score: 1

    make sure your up on the latest Client/Server technologies.

  86. Consultant by Troke · · Score: 1

    Find a consultant to help with both upgrade and initial maintenance, not only will they take a big load off of you initially for the proper setup, they can teach and show you what is applicable to YOUR network, and take the blame if things go wrong down the road. Full Disclosure: I'm a network consultant. PS. Do you want my card? I like the sounds of unlimited budget :P

  87. Dear Slashdot by hawguy · · Score: 2

    Dear Slashdot,

    After many years as an expert carpenter, I've found a need at my current employer for a plumber. I've made extensive use of plumbing in the past both for input and output and know I can handle the work. Many of the concepts are the same between carpentry and plumbing (i.e. cutting things and joining things), so I only need to brush up on the mechanics of how to do it. The pipes in our current building are all old and leaky, so we want to replace them. I have been given an 'unlimited' budget for pipes, tools, etc to set up a small toilet in the basement and after that I plan to replace all of our plumbing. What tools and materials should I acquire? What books should I read? Should I take classes?

  88. Re:Only 100 workstations by dstyle5 · · Score: 1

    The "Preview" button ate it all after you clicked on it. ;)

  89. To actually answer the question... by mengel · · Score: 1

    First: learn about networking generally. In your case I'd recommend the Doug Comer/Dave Stevens Xinu networking books, volumes I and II, but a lot of folks also like the books by W Richard Stevens TCP/IP Illustrated set. The Xinu books, particularly volume II, have the entire source code of a straightforward impelementation, which is really good if you're a person who reads code well.

    Then pick 2 network vendors you like and learn how to configure their gear. Probably start with whatever gear you have now; it may be perfectly serviceable if setup properly, or at least usable as a corner of a better network design.

    --
    - "History shows again and again how nature points out the folly of men" -- Blue Oyster Cult, 'Godzilla'
  90. PROTIP by zer0won · · Score: 1

    Google "networking." Seriously though, since "unlimited" budget in my experience usually means "nonexistant," I'd have to say pfSense and OpenVPN ftw. If they are that small there is no way they can really afford to give you a limitless budget. What exactly is running on these blades? Any specialized software? What kind of business is this? How is the wiring in the building? If you want to build a network for a business, you need to understand what they need to do and find the most transparent way to let them do that. You should also google "networking."

  91. Juniper? AHAHAHAHAHA by SuperBanana · · Score: 1

    Deploy Juniper products where you can. Commit confirmed alone will help keep you sane.

    You mean the firewall vendor that can't even get passive FTP right?

    http://www.google.com/advanced_search?q=juniper+FTP+ALG

  92. Some meta-advice for the newbie by taustin · · Score: 1

    I learned how to do this the hard way, by screwing up a lot. Here are some rules:

    #1: Always have a fallback plan. Assume that every thing you touch will literally explode in flames and kill everyone near by, and have a plan to return to the old stuff. It will cost more to do it that way, but being able to fall back to something that's worked for years will cost a lot less than the network being down for days while you have no idea how to fix it.

    #2: Test, test, test. And then, test some more. And assume that you have only tested 1/10th of what the users will actually do. Pick out two users to help you test: the best with computers, and the worst. Between them, they'll find the stupidest and smartest mistakes you miss.

    #3: Complain a lot. Seriously. Complain about what a pain in the ass stuff is to do stuff, and give examples. But show consistent progress. Nobody will have any clue what you're talking about, but they'll get the idea that you're working hard at it, and that you're succeeding. The biggest issued I've had was when major projects went off so smoothly (because I was well prepared) that nobody noticed the work that went it to it. After getting a COLA raise after a two year project to build a mission critical WAN, I learned.

  93. Buy Novell, yes, the whole Novell company by egork · · Score: 1

    You say you have unlimited budget - just buy Novell back from Attachmate. There will certainly be one or two knowledgeable guy who could run your network for you.
    For giving you such an amazing advice for free, may I suggest you keep Novell suing SCO out of their pants?

  94. We all learn by trying.. that's what classes do. by Weezul · · Score: 1

    Classes are great for teaching "theory" to "practical people" because the "theory oriented guy" teaching the class understands what "trying things" means in his particular domain. I'd imagine you've used some calculous for work on occasion, but presumably you'd never have sat down and just tried things involving integrals.

    I'd expect all those system administration courses are designed for people who don't know soo much about computers, but need some elementary networking theory before they'll become remotely competent administrators. If this guy's a developer, then he's likely already seen anything those students would find "theoretical", meaning he's already well set up for "just trying" more practical stuff.

    --
    The Christian religion has been and still is the principal enemy of moral progress in the world. -- Bertrand Russell
  95. A system Admin by JumpDrive · · Score: 1

    Someone has already suggested Tannenbaum's book. "Computer Networks"

    I worked for 3 years in a MS shop supporting MS systems for clients, I attended classes and earned my MSCE. All of the classes which I took for MS were centered around 'go to this menu' 'click this' 'entered this'. Yes I could set up networks, VPN's, servers, but I didn't really start learning the nuts and bolts until I started working for a company that supported Linux servers.

    Once you understand the basics from Tannenbaum's book then you might go to 'Wireshark Network Analysis' by Laura Chappell.

    If you have time go to 'Linux Firewalls' I used Ziegler's book, but there are a number of sources. On your network you may want to try and setup a muilti-homed linux system protecting a linux system. You can learn a lot about how a computer handles packets by setting up a server with iptables and adjusting it to allow certain packets or to allow only certain services. Also look at studying Computer security and security tools.

    Switches, You should look for switches that will allow you to monitor ports on the switch. If you only have a 100 computers on your network they may not allow you to budget for a network analyst to come in and analyze your network if there is a problem.

    Look for any tools, such as Snort or installing a computer which can tap into your internet traffic, that will give you a raw eye onto what is happening on the network. Nothing is worse than having a network problem with multiple computers on the network and being blind.

    You will still get to use your programming skills, there are few programs which offer exactly what you want or need to monitor your network and your computer systems. Eventually once you know more about what you want, then you can take results from other programs and craft your own system with reporting tools.

    To prepare you for this, imagine you have just been given a new position within the company where the previous programmer has left the company. You have an application which he has built up and put into production over the last five years. You are now in charge and there are no comments on the code. Your first ethernet card failure will be like your first bug. Same for switches. No one told you they could fail like that. Welcome to Hell.

  96. Man-up by Espen · · Score: 1

    Seriously? "Maybe 100 machines" and people here are advising you to run for the hills? Can I send this: Diagnosis to all of you? (Don't worry, it's not going to bring the internet to it's knees).

  97. I feel for you... by surfcow · · Score: 1

    I hope you *insisted* on getting some kind of compensation for OT.
    The job will eat your life - if you let it.

    Read:
    TCP/IP Network Administration - Hunt
    Essential System Administration - Frisch
    The Practice of System and Network Administration - Limoncelli

    Is the current system seriously broken?
    Did the previous admin leave any useful documentation?

  98. Welcome to The Suck, except it pays well... by JakFrost · · Score: 1

    Learn the difference between Support (Administration) and Development (Programming). The best you can do in Administration is put things back they way there were before they all broke so the users stop complaining or at best make small and slight improvements to a screwed up systm. Your job as administrator is to deal always with problems and very little time is devoted to improvements. In Programming you fix or improve applications or build fresh new ones. Go back to Develpment if you still have a choice.

    Study

    Otherwise if this is a bottom-up approach to learning networking and server administration without any previous hands-on experience with servers and managed switches then start with the certification tracks and books because they are well planned out, have plenty of books available, have training classes or web instructions. Get books and materials for CompTIA A+, Network+, Server+, Inet+, Security+, Linux+ since you can learn these generalized topics quickly and easily and at least be introduced to very basic ideas and terms that you never had to deal with such as RAID, iSCSI, LUN, VLANs, trunks, aggregated links, routing, CIDR, OSPF, core, edge, etc. You can skip the CompTIA tests for those since they are not really worth the money anymore with expiration dates.

    Move onto Cisco CCNA for more in-depth networking and one of the CCNP specialties for advanced topics. Touch some Microsoft and Linux server admin certs because you'll need to understand the servers and how they actually function and comminicate on the network to plan out your setup.

    If you have a chance look into SANs (storage area networks) and expecially iSCSI (i.e. Storage over Ethernet) because you will have to support it now or very shortly. Fibre Channel also while you're at it, learn zoning, provisioning, find out who Brocade, Qlogic, and EMC are.

    Also be sure to learn about Wireless networking using Enterprise level products and access points because that can get difficult quickly when it comes to proper setup, authentication, RADIUS, encryption, WPA2, TKIP/AES, certificates and auto-enrollment.

    Vendors

    Blades - HP, Dell, IBM, Cisco
    Switches - HP ProCurve, Juniper, Cisco, Netgear, Alcatel
    Servers - HP, Dell, IBM

    Recommendations

    Network - 10.A.B.C/8 for your network. A=site or core segments, B.=floor or edge, C=each subnet. Use /24 as default mask for 254 hosts and /22 for larger special subnets for terminal servers, virtual guest farms.
    DMZ - NAT your public IPs to a dedicated DMZ VLAN, firewall it from inside with static explicit per IP and Port rules.
    Vendor Systems - Segregate on seperate VLANs or subnets since you don't own or control these devices. Firewall from Production.

    Production Network - Keep small /24 subnets, edge to core (i.e. like a pointed star with center as core or multipe stars joined at core)
    Workstation Network - VLAN and keep subnets small and logically devided by physical barriers, floor, building, site, etc.
    Server Network - Try to keep server types separated on their own VLANs and subnets and concentrate them physically and by switch/card. Separate unlike and strage servers, applicances, vendor boxes away from regulat servers.
    Backup Network - Physicially separate the cables, switches/cards for workstation and server centralized backups, (Symantec/Veritas NetBackup).
    Wireless Network - Firewall and separate on VLAN
    Virtual Server Network - Dedicated VLAN and get 10Gb cards for switches and servers/blades.
    Storage Network (iSCSI) - Dedicated cables, switches/cards.
    VoIP Network - Separate VLANs & inter-switch trunks to keep away from all other traffic, separate switches/cards for sure.

    There are many more suggestions but at this point I'd have to start charging consulting fees. Find people to help you and pay them well.

    Good luck!

  99. Only a 100 devices? Cake. by MoldySpore · · Score: 1

    That is easy.

    1) Buy 3x 48 port Cisco 2960's will give you more than enough ports with some expandability (3 x 48 = 144 - 3 (for uplinks) = ~141 ports for devices) at a decent price, especially if you can get on some kind of discount contract (such as state price, etc.). Layer 3 switches are the way to go. Configure the switches inside a single vlan (unless you want to get fancy).

    2) As long as you are getting an Ethernet hand-off from whatever ISP you are using (fiber seems unnecessary for 100 devices), get a Cisco ASA 5505 with the UNLIMITED license (ASA5505-UL-BUN-K9) so you won't have to deal with user licenses or if the network grows past the limited licenses. 5505's are perfect because they are super cheap and provide most of the functionality of a higher end model like a 5520 but are mostly just lacking GBIC slots for fiber handoffs. The ASA will act as your firewall, and allow for remote connections using VPN (using Cisco's awesome AnyConnect client). Follow online instructions for getting ASDM setup for the ASA so you don't have to deal with CLI and can do straight GUI configuration (very handy if you don't feel like learning to program an ASA from the command line).

    3) Configure a simple network (with only 100 devices you can get away with a /24 mask on whatever subnet you use (255.255.255.0 mask). If you want you can go 255.255.254.0 and give yourself some extra breathing room if you think the network will grow past 254 unique devices. Configure your DHCP server (or whatever will hand out addresses for DHCP) to leave a range for static IP's that you will set on your servers. Workstations can pull DHCP as long as you have an internal DNS server so that people can remote to their desktops via the computer name. If not, then you'll need statics on your workstations as well for remote desktop.

    4) Depending on your needs, you can add a few wireless access points to the mix as well to blanket the area in wireless. Preferably I like to use a controller (I use a 5508 @ work) but that might be overkill for you since you'd most likely only need a handful of access points. Although a 4400 with support for a limited number of AP's would be nice and on an "unlimited" budget, managing it is cake with the controller. Anyway, get the CIsco 1142's, esp if you go the controller route, since they can come with the LWAPP (Lightweight Access Point Protocol) enabled IOS image already on there. Don't forget to consider power/ethernet drops to where you will be placing the access points, and do a wireless survey with a test unit and a program like inSSIDer to gauge distances between where you should place them for maximum coverage.

    I am a network engineer for my day job, and 1 of only 2 people who manage and maintain an enterprise network of over 5000 devices and ~8000 users. While we are stretched pretty thin, we manage to take care of that account and still have time to do things on other contracts (we are contractors), although there is a separate IT help desk staff at the main location to deal with specific user issues and workstation stuff so at least we aren't removing viruses and crap like that.

    I did go to school for it (Bachelor's in Network Engineering) and got certifications, but really unless its a huge enterprise network there isn't much of a learning curve beyond the CLI commands and maybe wrapping your head around some ASA/Firewall stuff (NATs and Access Lists will be your main nemesis). Especially if they are giving you an unlimited budget, you can take a few classes to get the basics down ("Networking Fundamentals") and go from there if you are more comfortable having some kind of founcation. The biggest thing is getting used to the syntax of programming a device via CLI, and if you were a programmer you will most likely pick it up fast. If you can get your hands on a couple switches to play with, between that and Google you will be able to setup a fairly simple network to provide all the services I menti

    --

    "I hope you know how very lucky you are to know me, because I am so incredibly incredible."

    1. Re:Only a 100 devices? Cake. by Vrtigo1 · · Score: 1

      2960s are layer 2 switches - you say that layer 3 is the way to go but give a layer 2 switch as an example which might be confusing. For 100 devices, there is really no need for a layer 3 switch unless you have a need to segregate devices into different VLANs and need wire speed routing between the VLANs.

    2. Re:Only a 100 devices? Cake. by MoldySpore · · Score: 1

      Whoops yea I should have included some kind of aggregate layer 3 device like a 3750 or a similar device that all the 2960's collapse into. Read it 3 times and still missed something ;)

      --

      "I hope you know how very lucky you are to know me, because I am so incredibly incredible."

    3. Re:Only a 100 devices? Cake. by mjwx · · Score: 1

      (3 x 48 = 144 - 3 (for uplinks) = ~141 ports for devices)

      You'll need a minimum of 4 ports for uplinks if you're using Ethernet. Plus one more for the link to the WAN connection.

      Layer 3 switches are the way to go

      Layer 3 switches are routers, what you need are layer 2 switches with management, but at 100 end points this will not be necessary, the only time you'd want management is if you had to VLAN off a VoIP network. but at 100 end points, unless you're running a call centre gigabit Ethernet should be able to hack it.

      get a Cisco ASA 5505

      Forget this, unless you're a cisco engineer most of it will be lost on you.

      Get something simple like a Fortigate 60 and forget about Cisco's licensing model.

      Cisco really is not built for a network of only 100 users. 1000 users perhaps, but not 100. The 48 port Cisco switches will be fine, but other brands will be just as good.

      --
      Calling someone a "hater" only means you can not rationally rebut their argument.
    4. Re:Only a 100 devices? Cake. by MoldySpore · · Score: 1

      Actually, layer 3 switches are layer 3 switches, not routers. True routers (such as the Cisco 2600 series) are true routers. It's rare to see a true router used anymore in an enterprise network outside of simple VoIP setups running CME (Call manager Express) or the need for a T-card for T connection hand-offs. Other than that it can all be accomplished with a layer 3 switch. Layer 3 switches are almost always used for aggregate devices and routing purposes instead of routers. Also, preferably, you would want to VLAN off anything you can but in an environment with only 100 devices it is mostly unnecessary. A flat network would be much easier for them to setup. But when you build an enterprise level network, especially with an "unlimited" budget, you plan for future growth and you anticipate the future needs of the network. Having a layer 3 switch (such as a 3750) act as an aggregate device that the 2960's (which are layer 2 switches) collapse into, allows for future migration to a more heavily VLAN'd network, among other enhancements. Really, the wireless traffic and wired network should be inside their own vlan, especially if you plan of offering a public or guest SSID for non-company people to connect to. Also, I wouldn't recommend mixing vendor's on their first network they build. Different vendors can act weird with each other and have extra troubleshooting or setup steps not normally necessary if they'd stick with all Cisco or all of another brand, especially for the firewall/VPN device that will be controlling access in and out of the network.

      --

      "I hope you know how very lucky you are to know me, because I am so incredibly incredible."

  100. Falling Star? by skartek · · Score: 1

    This must be a hoax message - either way the post and subsequent comments have made my day. Welcome to hell.

  101. Re:Network+ by rcoxdav · · Score: 1

    I would tend to agree with that, except with the subnetting. I have seen a lot of people have problems with it. Of course, being a programmer gives you a leg up, as you have probably dealt with binary and Boolean before also.

  102. Go with whatever works the best by FridayBob · · Score: 1

    If you do you'll always end up with FOSS (Free and Open Source Software) solutions and love your job. Either that or you'll end up with too much stuff that doesn't work and you'll hate your job.

    That's the problem with this line of work. More than anything, sysadmins love stuff that works and it's often the grubby little details that make all the difference. However, it usually doesn't work out this way for them. If you're serious about your job, it won't take long before you to realize two important things about the world you live in:

    * Commercial software companies just want your money
    Unfortunately, it's not just about selling licenses (if only that were true): it's also about limiting the customers in what they can do with the products after they buy them. For instance, they typically use proprietary file formats, databases and protocols so that their products do not work with those of anyone else except their own, or those of their choosing. They call this "being competitive", but it's simply about limiting your choices so that they can steer you (the consumer) in whatever direction they want. They don't care about helping you to get everything to work they way you'd like, so functionality is severely limited. Basically, these companies just want your money, and preferably as much of it as they can get.

    * FOSS developers just want things to work
    These are the people who are on your side. They're just like you: they dream of systems that do everything you want and of users who are happy. Your system should be like that box of Lego blocks that you played with when you were a kid: everything fits together, even if now things are much more complex. This is accomplished by using open standards for file formats, databases and communications, as well as by providing the source code for the software.

    What this means is that your success and happiness will depend on how much you can limit the use of commercial software in your network environment. Unfortunately, the average user (including your boss) has zero understanding of these concepts. Instead, their choices are much more likely to be influenced by a complex combination psychological factors, such as the marketing efforts of the big commercial software companies, their own limited experience, the advice of their favorite vendors and sales representatives, their desire to avoid learning to work with anything new, and even what their friends think they should do. In other words, unless they really respect you, your advice will not be taken seriously. Instead, they will likely tell you what software to work with and your efforts will ultimately be frustrating, the results disappointing. However, if you're lucky and good at explaining, maybe your boss won't blame you too much for the results.

  103. It happened to me over a decade ago by Gim+Tom · · Score: 1

    Pretty much the same thing happened to me back in the late 1990's. However, I had a very good background in hardware (my first programming language was a soldering iron) and I am a graduate engineer. If you learn fast and don't believe too much of what sales weasels tell you then you might be all right. Being very comfortable with both hardware and software is a real advantage since it gives you a perspective from both sides of the great divide. Don't focus too much on any one vendors "solution d'jour" but try to understand the real principles behind what you are doing. The best example of that I can think of off hand is if you know what a Netmask REALLY does in the hardware and think of it in BINARY then you never have to memorize silly rules about how to figure out what one needs to be. Hint -- study the XOR function and realize that an XOR gate was and is a piece of hardware. It took me several months of intensive study (mostly nights and weekends) to get up to speed and I never really stopped learning. The only formal training I ever got was a three day class on the Firewall we were going to use and that was well worth it, since it got me started thinking about how EVERYTHING had an effect on Network Security -- It also got me another hat a few years latter as the Network Security Officer. Now, for the downside. Networks are utilities and the only time your work is REALLY noticed is when the network is down. As a network administrator no one notices if you do your job right -- outages just never happen. When things do happen, however, they are ALWAYS your fault and you ALWAYS take far too long to fix them -- even if it is 3am in a blizzard when you had to come in and do it! Good Luck

  104. You should get a lawnmower, chainsaw, truck first by Sulfate · · Score: 1

    So you want to be a sysad eh? Well there's only one good sysad and that's the bastard who has become one with the bastard.

    First your going to need a small wad of money, cause your going to want to own the network, not work for it.
    To get that wad, I suggest lawnmowers, chainsaws, trimmers, edgers, roto-tillers, backhoes, and trucks, don't laugh, just do it, knock door to door and get jobs until you make a name for yourself.
    Work and Try to get a General Contractors license.

    You have your wad of cash and you still want to be a sysad.
    No you don't. Not yet.

    Learn Lockpicking, Survival, First Aid, Electronics, Transmitters, Receivers, Gardening, Pipe Fitting, Drilling, Trenching, Soldering, Engine Maintenance, Electrical Power Distribution, Load Control, Basic Business and Law. It helps to know lots of math, lots of on the fly scripting, if you can program yourself out of any problem, your close to being ready.
    Your actually starting to live like a king now and you still want to be a sysad,
    But not yet.

    Start investing in precious metals like gold, silver, copper, learn to read world events. You'll need a pistol now.
    Start studying international corporations, how do they work, where does their money come and go to, what does their infrastructure look like, where are they located, is it close to your resources, or too costly to do a TDY? Sell some excess tools, tech, on eBay, take a vacation somewhere exotic. Make connections, friends, people you can sub-contract and trust to get things done because they're independent and fast.

    Learn photography, cameras, surveillance, audio mastering, streaming
    Make a few plans for networks. Make friends with Senators, Congress, Chamber of Commerce, pnac, aipac, cfr, rothchilds, bilderbergs, DHS, cops, and sheriff.

    When you finally do find a target, tear it up. Your ready for nearly anything.
    You could be the one who pulls a national psyop. A bastard you are now!

  105. Work hard. Verify everything. by dweller_below · · Score: 1

    I went through a similar process. You will only survive if you work hard.

    Start looking at packets now. You must eat, sleep, and breath packets to survive. Use Wireshark and TCPDump. Don't let anybody abstract away any of the layers. You have to understand every network layer from 1 to 4 before you can begin. You have to be able to think like a packet.

    Physically touch and diagram every piece of network equipment. You must be able to draw a map of your network from memory. DRAW the map, verify it's accuracy and keep it in a safe place. When something goes wrong, you will forget everything and that map will become very important to you.

    You can have reliability or complexity. You can't have both. Educate yourself, then educate your boss. Make sure he understands that any complexity will reduce reliability. If you can't agree on the level of reliability and complexity, find another job.

    Don't believe salesmen. Cisco sales are worst. They will destroy you in a minute if it means a sale. Divide all Cisco performance figures by 3 to get YOUR performance. At your size, you should be able to mostly avoid Cisco. Avoid them as much as possible. If your network design is simple, HP and Foundry (now Brocade) switches will consistently outperform Cisco, dollar for dollar.

    Don't believe vendor performance figures. Evaluate equipment based on your own measurements.

    READ THE BUGFIXES for the current and previous versions of your firmware. There are always more bugs. Future bugs will tend to occur in the same feature sets that gave rise to previous bugs.

    Wait till you have a year or two of experience before tackling the following feature sets:

    1) Redundancy. Redundancy is Cisco slang for: "I sold unnecessary equipment to a gullible customer." Redundancy is hard. In spite of everything you have heard, redundancy virtually always reduces reliability. Simple network designs, based on simple equipment will almost always be more reliable than redundant ones. Don't experiment with redundancy until you completely understand your network. Then only deploy redundancy after extensive testing.

    2) VLANs. VLANs are a simple idea that enable you to create limitless complexity. Once you start, you will not stop until you have created a network that you can not understand or debug.

    3) Multicast. You are not a true network person, until you loath and despise multicast. Wait till you fully understand why you hate multicast, before you depend on it.

    Surprisingly, you should not hesitate to play with IPv6 (in a non-production environment of course). Nobody else understands all the implications of IPv6. It is one of the few areas where you will not be at a disadvantage :)

    Miles

  106. Two Time CCIE reply by Anonymous Coward · · Score: 1

    Hi, My name is Peter Revill, I have a little blog you might like to visit at http://ccierants.blogspot.com
    I also have two CCIE's (CCIE #18371 Routing and Switching, CCIE #18371 Voice)

    First of all, I want to address the idea that there is no jobs for network adminstrators, recently the company i was working for previously had some difficulties and I left: I had more offers than I knew what to do with, I am not trying to brag, i am just trying to allay fears that there is no work. I took a voice bootcamp in San Jose and all my class mates are drowning in job offers.

    So there you have it, the job market is strong for network engineers, final proof would be

    http://www.itjobswatch.co.uk/

    Look up CCNA and CCIE etc on that, you will see a good demand

    so that's my first bit of advice, second bit of advice would be: Aim extremely high and keep going, everyone get's there CCNA and then settles, that's not enough.

    Finally: So much Cisco Bagging going on here, No idea why: the fact is that Cisco realise the network is more than just a way to connect PC's, it is a platform in and of itself, converged voice video and data (Unified Communications) is a great architecture and works very well. IP can transport _anything_ and we might as well start using it everywhere. Please take my advice on this: Cisco is not going anywhere, it's in the most demand of any vendor and will always win out over other vendors when it comes to features.

    I hope this helps

    1. Re:Two Time CCIE reply by 1s44c · · Score: 1

      IP can transport _anything_

      Actually it can't. It can only transport digital data, analog data can't be reduced to digital without loss.

    2. Re:Two Time CCIE reply by 1s44c · · Score: 1

      Actually it can't. It can only transport digital data, analog data can't be reduced to digital without loss.

      I'm not sure what you mean by "analog data," but for signals, you certainly can encode it digitally without loss, depending on your bandwidth and the noise floor of your analog system.

      That is to say, given a high-quality AD/DA converter and a sufficiently high resolution (sampling rate and quantization), you can encode and decode an analog signal and get a result indistinguishable from the original within the tolerances of your gear.

      Encoding analog to digital always gives an approximation of the real data, this is what you are refering to as 'indistinguishable'. It's not a copy of the real data, it's just close enough to look like a copy. For encoding sound intended for human ears it doesn't really matter, for encoding something else the difference may matter. There may be many things in the universe that can't be accurately described by a finite length string of ones and zeros, the obvious example being pi.

  107. You can't be serious? Anyway, read this book. by quaero_notitia · · Score: 1

    The Practice of System and Network Administration, 2nd Ed. by Limoncelli, Hogan, and Chalup. You've got so much to learn technically and administratively, but it can be done with time.

    --
    -- Wondering how long until the internet becomes fully corporatist, like television.
  108. Get Help by Vrtigo1 · · Score: 1

    If you're serious about the unlimited budget thing, and you want to be responsible, don't try to "learn by doing" with someone's production network. Go out and hire a network consultant to help you. A consultant shouldn't have any problem if you want to be the one pushing the buttons, but you should have someone there to at least check your work to verify that what you're doing is sane and isn't going to cause any type of major problems down the road. As far as learning, there is so much covered by "networking" that you really can't learn effectively by being thrust into a position where you have to come up with something that works on the first try. You need to identify the specific areas that are important to you. A good TCP/IP foundation is pretty much a universal requirement for anything to do with networking these days, so I would recommend either getting a good network essentials book or taking a class at your local CC. A lot of the early networking stuff is memorization, I.E. what's the max cable length for the various types of ethernet, what is an SSID, how do you update the code on a router, etc. Being totally realistic with you, a lot of the early on, basic stuff isn't that fun to learn about. Especially if you're coming from the programming discipline. You'll get bored very easily. But once you start doing more advanced stuff it gets more interesting. I've always enjoyed VPN and firewall/security stuff more than configuring routers. Once you learn about VLANs and the cool stuff you can do with them, you will start to see the various ways you can accomplish a task, and you'll see where you have the opportunity to get creative with your solutions. Virtualization and storage networking are huge right now, and there's a ton of good info you can learn about them available for free online. If you have a basic, flat network, then it's not all that complicated to get something up and running. Buy a firewall to do your NAT/VPN for you, connect that to switches for your clients and servers and you're off to the races. Cisco is a good recommendation and I'll tell you why. Cisco engineers are widely available everywhere, plus you have the TAC at your disposal. If you run in to a problem that you can't figure out, you have very well defined avenues for getting help. Juniper engineers are also out there, but they're not as easy to find, and that may command a price premium.

  109. With an unlimited budget... by Junior+J.+Junior+III · · Score: 1

    1. Take the money.
    2. Hire someone else to do it.
    3. Charge the company 2x what you're paying that guy. (PROFIT!)

    --
    You see? You see? Your stupid minds! Stupid! Stupid!
  110. Re:Work hard. Verify everything. Yea Verily by Gim+Tom · · Score: 1

    Truer words have not been spoken -- especially about Cisco Sales weasels. I had one tell me that I couldn't mix brands of Routers and Switches in a network. Specifically that the 3-Com switches we had then would NOT work with their 2600 series routers. They were wrong of course.

  111. As I've learned from Facebook by /dev/trash · · Score: 1

    Changing inode_bits to 64 from 32 may get ya better performance but it is a one way conversion AND nfs mounted clients don't convert 32 to 64.

  112. Turning a cat into a dog? by Scotty+L · · Score: 1

    Sys Admins and programmers are two completely different animals. When most of us were young we made a decision on which path to follow. The path of enlightenment, or the path of programming. If there is an unlimited budget, hire an admin.

  113. Fortigate. by mjwx · · Score: 1

    Dead simple installations, multitude of configuration options to do most everything. Still lets you get down and dirty if you need to.

    Unless you're trying to do something like server publishing or VPN, in which case it will fail repeatedly and you'll get no support from Sonicwall. Not to mention the slow, pants on head retarded UI and poorly written help files. Try a Foritgate instead. Fortinet publish comprehensive admin guides for their FortiOS and dedicated guides to connecting VPN. Easy to configure from scratch, have useful metrics and logging not to mention a command line built into the web based UI.

    --
    Calling someone a "hater" only means you can not rationally rebut their argument.
  114. Former GKN instructor butts in with more :) by LostMyBeaver · · Score: 1

    As a former instructor (way back in the previous millennium) at Global Knowledge, I found that the Cisco classes at Global Knowledge are fantastic. I "audited" several of them and learned tremendous amounts from them. But, let me be frank. This is 2011. Unless you need hand holding or a paid vacation, video courses tend to be equally complete (especially the ones which use the Cisco course materials) and are MUCH more affordable. The budget could be spent elsewhere.

    Also as a "star programmer", there are many questions you should have which could never be covered properly in a Cisco training course. I personally develop network infrastructure equipment for broadcast video over IP. As a programmer, when I hear about routing protocols, I want to clearly understand the protocols. Recently, I have been working on developing an in-house course for my colleagues regarding IPv6. Compared to the Cisco courses on the topic, it's far more compact and far more detailed. It makes the assumptions before starting that I don't need to explain hexadecimal, it's assumed that when discussing the routing protocol instead of spending ages covering drop through mechanisms, I can simply present the algorithm and it will be understood. So the Cisco theory classes are insanely overpriced and painfully boring for "star programmers".

    I'll chime in on #10 as well.

    Now that you're not being paid to program any more, take advantage of it. If you need a programmer for a task, hire one. Programming is something you should do for fun now. Find the open source project you always wanted to work on. And build on it. The hiring someone for a task theme is good for much of the rest of it as well. You can't be an expert on every technology. Hell, I want to be, but when it comes to the output quality of the video codec I'm working on, the girl with the Ph.D. that sits next to me is far better suited to design and prove those algorithms and I am. But I'm damn good at making them work and making suggestions as to where we should cut some corners if the math allows it. There are some things you'll have to maintain, but maintaining a mail server for example is just plain stupid. You can administer it, but outsource it if you can. Programmers have a bad tendency of reinventing the wheel over and over again. Remember, there's probably already a solution for it out there. Much of your job should be about finding the right one.

    An additional issue with issue 4. If you're not only the network administrator but also the help desk, use some of that unlimited budget to get a slave to answer support requests and reimage machines. Even if that slave barely knows a thing about computers, their your front line to make sure you're not spending your time changing toner cartridges. If you setup remote installation services (or whatever it's called this week) properly, it should be possible to train a monkey to install new images on machines. In fact, I once configured a system so that GRUB was installed on a USB stick (mounted within each computer) with imaging software on the stick. It made it so that you could reinstall the image yourself by pressing 5 keys on boot-up. If the users keep their documents on the network, this is a perfect solution. Then just keep your images up to date.

    The real point here, leaving technical details aside is that you shouldn't be imaging machines. I would however highly recommend you learn all you can about Windows PE and imaging services. StarGoat mentions you should buy the same machine in batches, but understand that with Windows 7, the same hardware imaging requirement is gone. Hard drive controllers are now standardized, you don't need a new driver for every machine, just a new optimal driver which will most likely be automatically installed from Windows Update. A well maintained image for desktop PCs is the key to a fluffy life.

    Also, this is 2011, you can bully users into using online services for most things. So, you can use Citrix or some other type of remote solution for application streaming.

  115. Damn ... by rrey · · Score: 1

    How can I filter all useless comments and leave only the comments answering the damn question ...

    1. Re:Damn ... by egork · · Score: 1

      ask a credible question in the first place?

  116. CCNA + TLDP by timtim1 · · Score: 1

    Definitely do CCNA. You need that. Then read all the Linux HOWTOs @ tldp.org That last one is maybe not apparent how its pertinent, but I have learnt things in there that I haven't learnt from any other book or teacher in my 8 years as an SA.

  117. What is the network used for by bWareiWare.co.uk · · Score: 1

    If the network is email/web for 100 non-technical office workers then just buy five £100 24-port switches and forget about it. If you are supporting 100 stock traders, or the storage/rendering for 100 3d/video editors then just hire five £100k pa network administrators and forget about it.

  118. You can't be serious by 1s44c · · Score: 1

    They gave you a network manager position when you don't know ANY networking? Who is running this company? A 5 year old child?

    Send me the name of your company, I have feng shui consulting to sell them...

  119. It's only 100 machines by dbIII · · Score: 1

    The article poster can do it with just about anything without a lot of trouble.

  120. They are stuck in the past by dbIII · · Score: 1

    Cheap whitebox gigabit switches from two or three dozen brands crap all over the stuff HP hasn't updated properly for years unless you want to also use the switches as firewalls - and even then there is midrange stuff that craps all over HP in both speed and features.

    1. Re:They are stuck in the past by Vancorps · · Score: 1

      I'd be curious what whitebox product can match or beat the speed of a midrange range or even high end ProCurve. Most companies out their overprovision capacity while a 48 port ProCurve will actually offer 48gigabits of throughput. Even on the low-end HP provisions like this but you lose a lot of what makes a managed switch great. It was one of the original selling points against Cisco since they never liked you filling up their switches with actual connections.

  121. "Star programmer" to "network monkey?" by EmagGeek · · Score: 1

    What the hell did you do to warrant such a demotion?

  122. Re:Only 100 workstations by dbIII · · Score: 1

    You've just reminded me why I'm so happy I run a network with mostly *nix machines on it and the people using it mostly act like adults. If I had to go around chasing malware I think I wouldn't be able to handle 1/4 the number of machines on my own.

  123. Hire an specialist. by Sait-kun · · Score: 1

    Promote yourself to the ICT manager.

    Hire an System Administrator experienced with Citrix.
    Hire an Network Administrator experienced with Cisco
    Hire someone for first line Servicedesk
    Hire someone for second line servicedesk

    That will be your normal every day in house staff.

    For the initial setup hire an company with experience in Citrix and with them build the new network. Make sure your own Sys and network admin work closely with the company so that knew exactly how everything works. Also make sure EVERYTHING is documented.

  124. 'Star' programmer?!? I've seen this before... by gsegelk · · Score: 1

    From my experience, someone who boasts to be a 'star' programmer usually is not one. I'm just saying...

  125. Re:Juniper? AHAHAHAHAHA by Slashdot+Parent · · Score: 1

    FTP? What year is this? 1993?

    --
    They don't grade fathers, but if your daughter's a stripper, you fucked up. --Chris Rock
  126. You're fine... by mcescalante · · Score: 1

    You're a programmer - I would hope you understand a fair amount of networking, etc. You should be able to pick this up quickly. Get whatever hardware - don't buy low end but you don't need to be buying Cicso crap or something that is intended for tens of thousands of end users. Read a few books about whatever system you're going to implement (Are they running Winblows Server?), and about TCP/IP / networking and you'll pick it up in 2 minutes (bet you could read a book a day). Definitely avoid anything Windows and AD - the network is too small to even be worth dealing with how awful it is. Maybe use CentOS (to avoid paying for Redhat Enterprise, as long as you're confident in your ability to fix stuff).

  127. Different skillset by lasinge · · Score: 1

    Troubleshooting networks is not hard, it's the peopleshooting that presents the real problem.

    --
    you are in a twisty maze of different passages.
  128. Network Administrator by hotelogix · · Score: 1

    To Become network administrator you have to take certification from Microsoft and Cisco without it no company will take you seriously