D E M E N T E D P E N G U I N . N E T

May 9, 2008

Debian install on a Dell 1950 not working?

Filed under: Uncategorized — admin @ 4:36 pm

We’ve been building a new product over the last few months where I work, and we’ve been ordering a bunch of Xeon dual quad core Dell 1950’s to put the new project on. We are a little unusual in that we use Debian where I’m at – which doesn’t make a lot of sense to me since Debian is every bit as good as RedHat, maybe even easier to use, and it’s free. How often does anyone call RedHat support?? Forums and mailing lists are usually more useful. Anyway….

The installs have been going great using a Debian 4.0 Net Install cd. I did set up a PXE boot server off of my laptop but the Dell’s take some special modules to detect the hard drives and I didn’t want to dig around to find out what they were, re-make the boot files and what-not. Besides, it only takes 15 minutes a box to install so it’s not
really that much of a time saver in the long run.

Last week we got our final shipment of servers. I’m doing the installs on them now but seem to have a problem once
the install process goes out to the mirror to get the extra packages. We do a standard system install which requires 109 files in all. It gets done downloading the 109th file and then just stops. Strange. Usually after the third time of rebooting and restarting the install it will finally work mysteriously. This is obviously related to either
a controller or drive module, but I’m not exactly sure why. Maybe the controller bios was updated since the last shipment we had, or the hardware changed in some way. The fourth system I started building out of the eight servers just refused to make it past the package download. I tried probably 10 times before I gave up. I Googled the problem but didn’t find anyone else having the same issue. I know I could contact the news groups to find an answer (probably a driver disk would be involved I’m sure) but I just didn’t have the time to wait for an answer or a solution.

I hit Alt-F2 and activated the terminal. I was digging around but didn’t see anything that would indicate where the lock up was happening. Finally, I decided to kill one of the dpkg-preconfigure install processes, which caused the install to error of course (hit Alt-F1 to get back to the install screen), which allowed me to go back and try the mirror download again without restarting the whole install. When I selected to retry the mirror download it picked up from where it had hung previously and completed the install. Strange, but yay.

I guess the point of this post is to let anyone else out there having this issue know of a work around to the problem without having to track down extra drivers or other viable solutions. Just kill the install and try again! I’m sure this would work with other distros too – like RedHat, Fedora, Suse, etc.

One other thing worth noting, the interfaces are switched by the OS. I.e. the phsical network connection labelled eth0 is actually eth1 and vice-versa during the install, as well as post-install. I start out putting the network cable in port labelled eth1 since it ends up being eth0 anyway.

Yeah, yeah – I’ll try to find the real solution and post it too, but I just don’t have the time right now. Don’t hold your breath either 🙂

Powered by WordPress