What do I need to set up a box server?
 

My current project is creating a linux server using an old pentium.

I'd like to have it set up without a monitor, keyboard mouse, and instead, operate it remotely from the network. Of course, on installing the distro, I would use a mouse, keyboard, and monitor, but only temporarily.

So how would I be able to access files on my server and add others to it? On searching, I've heard many terms being thrown around such as VPN, SSH, SAMBA, FTP, and I'm not exactly sure what I'd need? I've heard it's really simple to do, but I can't really find any information on how to set it up and use it. This is supposed to be a small file server and I was also thinking of eventually (but not decided) using it to host a website that I built in dreamweaver.

How would I remotely access and control it as if I were running the server directly through other linux machines on the network?

-Thanks in Advance

If your trying to access the file system with windows you need samba.
always install SSH, this way you can administer the server from any ssh client on your network. FTP works great to if your using dreamweaver because of it's built in ftp system. I like dreamweaver, in fact dreamweaver, PhotoShop and testing is only reason I have a windows box still around.

you could also try to download a small linux distro which runs on cd.
There are various small distro's which focus especially on security or functioning as server. especially the small distributions of linux are mostly very stable and fast on older machines.
please have a look here, maybe you'll find what you need.

http://www.distrowatch.com/dwres.php?resource=cd

There are some distro's which are quite small.. for instance: Devil Linux:

Devil-Linux was created by Heiko Zuerker, a US-based German IT manager (see this interview published by Portazero).

It is a specialist Linux distribution designed primarily to be used as a CD-based firewall or router.

The goal of Devil-Linux is to have a small, customizable and secure Linux installation. Of course, the main advantage of a CD-based firewall is that the files on the CD cannot be modified by an intruder - if you think somebody managed to break into your system, just reboot and the data in the RAM-disk are clean again. The other advantage is that it only takes 15 - 30 minutes to have your firewall up and running.

Good idea, Jan, but eventually I'd like it to be a little customizable. Although I might have a use for it before I get my box setup. Soon as I figure it out :cool:

Micxz,

I love dreamweaver because of its relative simplicity and overall..well you know ;)

Could you explain the process to me, though? I have zero experience running any sort of server. From what I understand now, I install all of those programs, and turn the machine on, but beyond that I have no clue as to how I would use it. I'm running through a router and it will eventuall have it's own static IP of 192.168.1.27 through gateway 192.168.1.1. So what process would I follow to upload to the ftp? Where would the files be stored (in /-- root I presume?) And if I ever wanted to run my webpage from it, how would I set that up?

Thank you in advance for any answers you can give for my (many) questions. I really don't know the last thing about what I'm doing. Only that it's possible.

Please?

First thing is to make sure that your computer will let you boot without a keyboard. Several older Compaq's (for instance) will not run headless.

Keyboard error
Press F1 to continue

(still the silliest error message ever)

Second - make sure you install a relatively easy-to-use distribution. Search here for MANY heated debates on easy to use. My humble option is RedHat 8 or 9 because it's almost as easy to install as Mandrake, and a large majority of downloadable binary packages are built for RedHat.

Third - when installing the distribution - install everything - especially the development stuff. You will regret it if you don't - compiling packages without them is near impossible.

Once your distribution is doing approximately what you want it to do - make an ssh connection to the box - unhook the keyboard - and use the ssh to reboot.

# shutdown -r now

Make sure it comes back up and that you can still ssh to it before taking the monitor and mouse off. Keep in mind once the mouse is gone - X may not run either.

I don't know if that's enough detail - but it should certainly give you some direction.

Thanks. :cool: I know I'll get this soon enough. Why? Because when I first attempted to install mandrake, I was on the edge of my seat. I'm experiencing that same feeling right now and so, things should work out with the server.

If you do find that your system will not run headless, one solution is to get a 2-port KVM (keyboard video mouse) switch and connect to it. These switches emulate presence of the peripherals at boot time, thus bypassing the problem. Both my Linux boxes run via KVM switches, with the other side of each switch going to a Win98SE box. Most of the time, the actual console is connected to the Win box since I still have to use these for my commercial work. However the Linux console is just a couple of keystrokes away...

How's it working out? You have linux installed right? Can you connect to the internet through the linux box?

It's working great; been using it for almost a year now and once I got through the initial learning curve it's been almost trouble-free. Haven't had to reboot the main box for almost four months, and that might not have been necessary but it seemed the simplest way out of trouble at the time.

One of my Linux boxes serves as the firewall and router for my entire system, and in addition provides a private FTP server, a web server, a local mail server, and a file server for the four Win98 boxes. The other Linux box currently just provides a caching DNS server, freeing me from dependence on my ISP's name servers and greatly speeding my browsing. You can try the web server as http://www.jimkyle.dns2go.com/ and see how it works; it's not yet fully functional though...

Both Linux boxes run Mandrake 8.1; I couldn't get the newer versions to do what I wanted, so rolled back to that release...

Well that's great Jim, But what I meant those questions to the starter of this thread, "flamesrock" to help him in his endevors. Me and you have our networks configured correctly but the point of this post is to help "flamesrock".

Thanks :cool:

I haven't even bought the box yet, but I know exactly what I'm getting (a small pentium2) and also what I want to install -- slack 9.1, probably without the gui. Everything, such as adding files and configuring will be done through NFS on a different computer(If I'm correct in my understanding of that). Although I might break down and install kde.

What I'm trying to figure out is how I'd set this up. I know that it would go through the router, and also that I want it to act as an FTP/webserver. So to do that I'd need:

Apache
Wu-ftpd
proftpd
ssh (I have no idea what this does)
and various other programs.

The problem is that I've never had any experience running those applications. To test, I thought I'd make a mandrake server partition on my primary computer. Apparently, it has the proper tools installed, but I can't figure out where to start to get the FTP up and also to start running the webserver for my site. This will be a very similar configuration to the box once it's set up (except a slightly different IP)

I have a .tk domain name..my GA is 192.168.1.1 and the IP is 192.168.1.25. NS 255.255.255.0 .

So how would I start using it as an FTP server and a webserver?

-Thanks in advance

I'd recommend you use proftpd rather than wu-ftpd; it's much more secure, and easier to configure. The sample configuration file that's part of the package will work nicely to get you started. Set it up to run standalone, though, rather than via inetd/xinetd. It doesn't take up much in the way of resources to do this, and avoids the slowdown of loading the program each time someone tries to connect.

My multi-server box is an old Pentium II, 200 MHx MMX, with only 80 MB of RAM in it. It does have two 40-GB drives, though.

For a window manager I use IceWM; I found KDE to be too bloated for my tastes on such a small system. I boot to the command line, then use "startx" to launch IceWM so that I can have an attractive screen saver running...

SSH is "secure shell" and its purpose is to let you administer the machine remotely. I don't use it; in fact I've disabled all the remote login services for security's sake. I do, however, have WebMin installed, and have restricted access to it so that only my LAN machines can get to it. I also have Samba installed. Between them they let me do most all admin tasks from any box on the LAN, Win98 or Linux. If you use a KVM switch you can run the server essentially headless most of the time yet still get into it via the KVM if need be -- and in that case you would not need SSH at all...

Thanks -- I think I understand SSH, now. But whats the difference between that and NFS?

I'm going to try IceWM on my mandrake test now. Sounds cool.

Sorry, I don't know the difference from NFS since I've never tried to use NFS (or SSH either for that matter). Once I managed to get Samba up and running I didn't need anything else for connectivity among all the systems on my LAN, and I definitely don't want any access to the admin capabilities from outside the LAN.

To get the screensaver you have to run xscreensaver as a background process. It includes dozens of different savers; my favorites are "pyro" and a space-battle thing...