NASLite Network Attached Storage

Let me try again then.

The facts you provide are:

1. You have an HP Vectra with a DVD attached to the onboard IDE.
2. You have an additional controller in that Vectra with a 500G SATA attached to it.
3. You have a good NASLite CD that boots fine on another machine.
4. You are able to boot a Fedora CD as well as other bootables on your Vectra as is.

The facts that I have to contribute:

1. I can confirm that a stock HP Vectra can boot, install and run NASLite from the CD as is.
2. I have no access to any data other than your case, that will indicate a software deficiency with the existing boot mechanism.
3. I have a pretty good grasp of how NASLite and it's associated utilities function.
4. NASLite uses a stock and unmodified implementation of the syslinux boot loader and in the case of the installer CD, the boot is El Torito while the Fedora uses isolinux, a different boot process altogether.

Considering the above I'm sure we can both make objective conclusions, I'd suggest the following as a start:

1. Remove all peripherals such as the Promise controller and try booting again.
2. Try replacing the existing boot DVD with another CD/DVD drive for the purpose of installing and/or testing.
3. Burn another CD at the slowest speed you can and try using that. Sometime fast burns work fine on the drive they were made on but do poorly elsewhere.

The only conclusion I can derive from the above info is that your particular hardware implementation has a problem with El Torito or your DVD drive is unable to load the complete boot.img file into RAM. BTW, the boot.img file can be imaged to a floppy and is completely open for you to examine if that will benefit your research. Like I pointed out above, stock syslinux is used throughout so you have unrestricted access to the boot code. You can find everything you need at http://syslinux.zytor.com/ and while you are there you can also take a look at the "hardware compatibility" section. I'll quote the first entry from that list:



Hope you find this more helpful than my previous posts.

"dim, I don't know what dimension you're in but you haven't contributed anything useful here yet. "

Yeah, ok dude.

Tremendously so Tony and I do appreciate your patience. You've provided plenty of meat in this post and I'll go offline and chew it. I'd considered interaction problems between the HP BIOS and the Promise BIOS and that may well be what is hanging the Naslite installer. The HP's BIOS seems to be completely unaware of the presence of the SATA drive in setup although the Promise controller does successfully hook the HP's BIOS during boot.

I've thought of one stupid trick of my own that I'm going to try on this machine to satisfy boot.img which is to enable the floppy controller and insert a blank floppy to ensure boot.img sees some local storage. If that doesn't work, I'll fully pursue all your suggestions and let you know the outcome.
Thanks for hanging in there,
Chris
P.S. I always burn bootable images at slow speed. I've been down that road before too.

PROGRESS!

Floppy stupid trick - No Joy
Remove all peripherals - No Joy
Swap CDROM drive - No Joy

If all else fails, surrender. Put everything back as it was and
temporarily connect IDE PATA drive - Success! Naslite CD boots.
Now Naslite can see both Disk-0 and Disk-1 so I select Disk-1 for install and voila it's done. Powered down, unplugged the temporary drive, used BIOS setup to set the first HD to "None" and rebooted. Naslite successfully boots from the SATA drive (which it now calls Disk-0). The rest of setup, registration, etc. HTML status looks good.

Some notes:
I was mistaken about the HP BIOS setup. It can see the Promise controller but it will not let the SATA drive be in the boot priority list unless all unused IDE HD slots are marked "None".
It also throws a POST beep code when no HDs are attached to its internal controller but proceeds with the boot anyway.
The HP Vectra has an internal cover switch. When you boot with the cover off it takes forever while it tests all RAM. Put the cover back on and it boots much faster.
I'm still curious why the Linux System Rescue CD's syslinux boots without a PATA drive while Naslite's syslinux won't. I'll try to dig into that one some more. Do you know which version of syslinux Naslite uses?
I'm also curious if I should want to swap the Promise controller for say an Adaptec, would my Naslite registration still be good? Right now the Promise seems to be working fine but experience may tell a different story.

Cheers!

Tony wrote somewhere else:

Some BIOS allows for legacy support where SATA can be allocated as IDE. You may try that approach or use an IDE drive

This little cheating worked when my DVD drive wouldn't mount. But of course this was a different board.

Yes Pieter, Tony was absolutely right about that, but unfortunately this is a legacy Pentium II motherboard made before SATA existed. The writers of its BIOS were forward looking and made provision for hard drives up to 32Gb (the originals were 4 or 6Gb) but they couldn't imagine the 500Gb SATA that it is driving now. It is a bit over engineered in terms of quality, but that used to be a hallmark of Hewlett-Packard.

Chris wrote:
'but they couldn't imagine the 500Gb SATA that it is driving now'

It must have been somewhere around 1987, that I bought my first -external- hard drive. My computer was an Atari 1040STf, and was basically operated with the modern single sided 3,5" floppies: booting, programs, storage. Then there was this giant hard drive of 20 Mb: an unattractive grey box, costing a small fortune. Boy, was I happy!

Those were great times and I can well remember the feeling of accomplishment with getting something state of the art working. I started a little earlier with a homebuilt system based on a Motorola 6800 cpu, but a "B" suffix that ran at a blazing 2Mhz! I had to write my own BIOS extensions and device drivers in assembler which were stored in 2K eproms that I burned myself. Originally, storage was cassette tape and ram was a massive 64K. Eventually I acquired a couple of IBM double sided, double density 8 inch floppy drives which were each capable of holding an incredible 2 Mb of data. I had to build my own floppy controller and the difficulty of reliably clocking data on and off a floppy at 4Mhz seemed almost unsurmountable. But I eventually figured it out with help from some trade magazines and had myself a real computer.

Around that time, advances like dynamic memory and colored graphic controllers started showing up so other than an upgrade to a 6809, my interest in building waned in favor of buying. I still build computers but that has a very different meaning now.