An argument for Linux in my workplace

[voidinit]

I work for a company that provides IT support for other companies too small to warrant their own IT staff. Recently I've come up with the following problem.

The Client:

The Phoenix Musuem of History is one of our clients. This musuem is dirt poor as far as IT budgets go, which is a shame because it's contact data is invaluble to it's operation. Currently it has Windows 2K server domain controller, DNS, file server and DHCP server all on the same box. This box is dying! It's old, broken and slow. The server reboots nightly and still has problems performing it's duties. Most importantly, the backups to DDS tape are failing consistently! This means we have an organization that is dependent upon data that cannot be reproduced if nescessary.

The Solution:

My organization has offered to replace their server with a used Dell PowerEdge complete with Windows Small Business Server 2003, Veritas Backup Exec, a new DLT drive and labor for $2800. This seems like a good deal, but the customer cannot afford to figure this into their budget.

My Solution:

Give the customer the new PowerEdge minus the DLT drive, tapes and Backup Exec., ( A savings of about $1,000 not including labor.) Then, let me take their old box, which is functional, and install linux + Samba. Then in about an hour I can use cron and bash to script all the backups nescessary to this once non functional box. This solution will cost the customer an extra hour at $110 per, but save them cost of a DLT drive, tapes and Veritas software.

The argument:

Our entire office is very Windows heavy. I had to "sleep with the enemy" and get a MCP to work here! Clint, who is our lead network/systems administrator is very opposed to the idea. He doesn't even know what bash or a kernel is. We get along great, so there is no grudge match here. His argument is well founded. If I were to walk out the door tommorow there would be no one at my company that knows a thing about Unix/Linux or any nix! Therefore he cannot allow me to put this solution in place because it "can not be supported". My response is of course "Well, you are supposed to be a sys admin.....Learn the sys!"

The long term:

Currently I have my own little world in the office. A LAMP system for document sharing and revison control. A guide to replace WatchDog firewalls with iptables cheap boxes. And of course a JBoss/Linux J2EE server. All of these boxes do their job very, very well, but will presently never see production because of that same damned non-supported argument!

The Solution:

I need a argument that I can put in front of the owner of the company, who worships microsft to a degree, but money more.

[Alex Cavnar, aka alc6379]

Originally posted by voidinit


The Solution:

I need a argument that I can put in front of the owner of the company, who worships microsft to a degree, but money more.

There's your answer right there. Why should you be paying through the nose for a Microsoft solution when a cheap Linux solution would do? If money's what he's truly after, then he should be ultra-objective. He shouldn't look at it from a "we don't support it" perspective; instead he should look at it from a "could we support it?" angle. If he does look at it like that, then I'll be the only thing he'll find standing in the way of Linux adoption for this project is his own ignorance of Linux. (I don't mean ignorance in an offensive light, I mean he just doesn't know it) Most people want to stick to what they know-- that's what they're comfortable with. Some have the luxury to pay a little more to get that comfort. But, in a case like your museum, they don't seem to have that luxury.

If you're looking at it from a dollars and cents perspective, then Linux will be cheaper in the long run. All of the TCO people can come in and say whatever they want to, but it's been my experience that you can do more for less and with less than Linux. Even factoring in the time to train people to adminster a Linux box, things will even out in the long run. I'd be willing to bet, even, that a PowerEdge server would be overkill for a Linux system doing all of that. Sure, it would be great for Windows, but you might not even need that many resources to do those jobs under Linux. What are the specs on the new server, anyways?

You should do a TCO study on Linux vs. Windows for this particular project. From the sounds of it, Linux would be the better choice. It sounds like this particular box really could just be locked away in a closet somewhere, only to be bothered during backups and other administrative tasks. Once you figure those TCO numbers, show them to your boss. Slap him in the face with them-- if he's really into making money, he'll be stupid not to go with whatever gives the most value.

[EnigmaOne]

Now that you have a good answer, how about a question....

Sometimes you can play the MCP/MCSE cert to greater credibility--especially when you're advocating against the ms camp. Sometimes it's just better to not say anything at all--especially when they're entrenched against all reason.

Is Clint totally oppposed to being introduced to Linux in any way, shape or form? (Can he be brought over from the Dark Side?)

[nko]

It sounds like they're not one of the biggest clients you've got. I'd guess they're small enough to where you can be more personal, and you might be able to make yourself their support contact. With a business that size, they'll probably understand if you explain to them that you're giving them something different from what was failing them, but your shop doesn't support it. All they can do is request your help, but that's not 100% promised to cut the mustard every time. You know where I'm going with this.

That might get the client a little more excited about the idea, and might also appeal some to Clint. Totally can't blame Clint for not wanting to pop some unknown software off to a client; he probably likes having a job, and going with an unknown solution's deeeengerous :-).

If you can get some of Clint's time to sit down and introduce him to the system, just make sure it's a rather stripped system, so that showing him stuff like /etc is as little overwhelming as possible, and try to not to get in to the nitty gritty; using a Linux system, I think, requires a little more attention to detail than a Windows system, and Windows admins don't like it when you're explaining how simple it is to configure Apache (all ya do is go in to /etc/apache, or whatever directory it's in, because some distros do it different, then use vi to edit the document root, and BTW press INSERT to edit things in vi, then save with :wq after hitting ESC, and then run top to find which PID apache is, kill it, then run httpd, unless it's Debian, then do apache...). Just an example to remind how overwhelmed a point-'n'-click admin might find Linux.

[dboyer]

It might be too sudden to throw a linux box at this situation... HOWEVER....

If you have the time/patience, it may be worth drawing up a TCO for this specific case of windows vs. linux. If you show it to clint, there is a chance that he will see that you can produce a "equivalent" product (yeah, yeah) for cheaper... The net result is that it would be easier for your company to get clients... It would be also easier to RETAIN clients because MCSE monkeys are a dime-a-dozen, but it is a bit harder (and more expensive) to find linux admins...

Sit down and show him the positive benefits of using Linux or OSS products... For example: if you're setting up a small office with 5 or 6 workstations, and can save the company hundreds of dollars by going with OO.org (a "equivalent" product) you may be able to get that contract over competing companies are still pushing MS word...

Thats the sort of language he will probably understand... of course, i would be careful saying something like "if that server was runing linux, they would have never had to call us"...

that sort of thing is a turn off for a boss :-) (unless you have yearly fees for unlimited callouts or something, but it sounds like you get paid per visit/job)

[voidinit]

Hrm....I'm sorry, I may not have explained the situation correctly. I am not replacing their MS domain controller or any application or file servers with a Linux solution at all!!! I'm moving their existing Win2k setup to new hardware, and I want to put linux on the old hardware to simply execute scripts that copies data from the Win2k Domain Controller to itself.

That being said, Clint is really, really outright opposed to Linux. I think he bad experience with a red hat install, meaning he couldn't figure it out in 5 minutes or less :), and has never looked at it again. He and my boss are in the same boat, they know it's out there but it's seen as evil and hardly viable. Nevermind that the've seen what I can do with one less than 1GHz box.

I'm just going to keep pushing them and pushing them untill one day there will be a situation where only linux will do! I think if I can break the ice with one production stable box we will be good to go.

[hlrguy]

Originally posted by voidinit
That being said, Clint is really, really outright opposed to Linux.

People usually fear what they don't know. That said, slip a knoppix CD into his CD drive of his computer and ask him to reboot. Or, try out the new Mandrake Live CD with USB storage. I ordered my copy. Great for me, telecommuting and travelling a lot. When it arrives, I will be posting my experiences with it (which computers/lpatops it works on, etc).

Another alternative, grab an old computer you have lying around, create your solution that backs up a Windows box at your work, and demonstrate it. If it works, it is a simple case of changing hostnames when you move it to the client site.

Heck, another alternative. Find out how much the client pays in Office liscences (they must have a few I would think), introduce them to openoffice, and then they have the money to fix their problems.

hlrguy

[shakin]

With all due respect to the other responders, non of those ideas will convince anybody to use Linux in your situation (TCO, in particular, is only valuable on internal deployments).

What you need to do is show them how they can make more money by working with Linux sometimes. In your example, they can get some money from a client who otherwise will not be able to buy anything. In other situations, they can bid lower than a competitor by reusing a client's existing hardware for various network tasks -- DNS, file server, mail server, etc -- that can be performed by Linux on low end machines.

Broadening their technical knowledge is not a bad thing. If you quit, why can they not hire someone else who knows Linux? If it's part of their business -- which it will be if they make money doing it -- then Linux is a good investment. You can even take charge of planning a pilot test of Linux on a few client sites. If it works, your company may decide a small Linux department can make money in the future... and guess who would be the natural choice to lead such a department? You :)

[AnonyPuss]

Why not just start your own business? You seem to know what you want to do. You're looking for any way that would save the customer money. The company that you work for isn't looking at all available options. You're being held back because others in your company are not knowledgeable about options.

The people at the museum would probably be very interested in anything that works better than what they have and at the same time saves them money. However, I'm sure it wouldn't go to well if you went to them directly. Not sure how to handle your boss, but previous suggestions sound good.

You really should consider starting your own business.

[Alex Cavnar, aka alc6379]

Originally posted by voidinit
Hrm....I'm sorry, I may not have explained the situation correctly. I am not replacing their MS domain controller or any application or file servers with a Linux solution at all!!! I'm moving their existing Win2k setup to new hardware, and I want to put linux on the old hardware to simply execute scripts that copies data from the Win2k Domain Controller to itself.

That being said, Clint is really, really outright opposed to Linux. I think he bad experience with a red hat install, meaning he couldn't figure it out in 5 minutes or less :), and has never looked at it again. He and my boss are in the same boat, they know it's out there but it's seen as evil and hardly viable. Nevermind that the've seen what I can do with one less than 1GHz box.


Okay... I get you now. In this case, Linux should be an even better fit. I like hlrguy's suggestion about setting it up in-house and proving to Clint that it will work. Personally, I'm a BIG proponent on testing the snot out of something before deploying it for any client, anyways. But, if you start to do little things in-house with Linux boxes, get some of your coworkers other than Clint "into it", then you may start building a more solid case for Linux.

The more people who know how to use it in your shop, the better. The more comfortable you can make Clint with Linux, the better. Just using Linux to do such a niche task as being the backup server could be enough to get a proverbial foot in the door for more involved projects.

But, on a second thought, why not just blow away the system install on that old machine and start over fresh with Win2k? If you could, say, go with a less powerful "new" box, and just have the old one as ther backup server, Windows 2000 should be sufficiently stable enough to handle that. Or, even better, why not start with a fresh install on that old machine, have it do all of the domain logons, etc, and get a lesser-powered machine to just be the backup server?

In the end, if the boss ain't happy, you ain't happy.

[voidinit]

The main reason I can't blow away the existing Win2k box and reinstall everthing is that doing so won't solve anything. The problems they are having is primarily due to the lack of rescources on a current box. I guess when they decided to buy this server they didn't think about things like adding employees and applications and databases growing and such. The old hardware just won't handle it anymore.

I like the idea of starting my own company. I am actually doing just that in a part time mode. It's a lot harder than I thought it would be... I guess if I could get some clients, it would be a lot easier. :)

I plan on talking with the owner and trying to convince him that his company can make money by deploying and supporting linux. I plan on showing him that even though it is a smaller market, there is a lot less competition in that market. Now that big boys like IBM are backing linux, it may just make my task easier.

As for testing the backup server in house, well that's already in place and has been since about a week after I started working there. My LAMP box and my Workstation are backed up nightly to a share on our pdc that is backed up to tape. To this day, I've not missed a backup. However that fact and the fact that my LAMP box has over 80 days of uptime seems to be overlooked by the powers that be.

I'm going to give Clint shell access to my linux box and point him to the proper documentation. Maybe a trip through the linux filesystem and a little discovery of the vast amounts of documentation and help there is for linux will ease his mind.

Clint knows as well as I do that it's always the hardware and the markup thereof that makes it very hard to sell a system to a customer. All the customer sees are numbers, very high numbers. Clint hates this fact as he really likes to have viable solutions in place, solutions like hardware redundancy and UPSes. These solutions keep us from getting paged on a Saturday and that is a good thing. I think I can use this angle to show him that he can sell good hardware a bit cheaper by not needing the horsepower to run Windows Server 2003. I hope anyway.

[ECartman]

I would say simply this:

"Linux is only getting bigger and bigger. Those unwilling to learn and adapt will find themselves extinct."

[gommo]

I think Microsoft has a ways to go before they become 'extinct'.

Though I think I know what you mean.

[thaddaeus]

Like said before, start your own business, and if Clint dosn't understand linux, he should talk to me. I was 14 when i first started(by the way my gpa is 1.8)using the worlds greates os. Maybe he shouldn't be working there if he can't install and os?, especially one that installs like windows.

[AnonyPuss]

Definitely start your own business, even though it may take some time to see the fruits of your labor.

The company you're working for obviously isn't concerned with what's best for their customers' wallet, but is more concerned with what's easiest for themselves. This is a business model that is doomed if it continues down that path. A GOOD IT man/woman/person should look at all possible scenerios, considering what is best AND cost-effective for their customer, and give their customer different options and prices.

Time consistently shows that those who do not adapt get left behind and die. It may take time, but they will die.

[voidinit]

Yes, I've started my own business and I'm only working at the place of my former employment full to three-quarter time. I don't have any clients yet, but my long time friend and fellow Marine Kyle may have a solution. He has started a web hosting/IT company and he already has a BSD guy, a windows guy (himself), he just needs a linux partner. I guess I'll be that partner.

[AnonyPuss]

Well, I don't believe in luck, but...Good Luck. I wish you well.

[voidinit]

Thanks, I think I'll need it.

[ECartman]

Originally posted by gommo
I think Microsoft has a ways to go before they become 'extinct'.

Though I think I know what you mean.

I think you severely misunderstood what I was saying. I meant IT "profesionals" unwilling to learn and adapt will find themselves out of a job. This can go either way too, Linux people unwilling to learn Windows may face the same fate.