I am working on decommissioning an old server for a client. The server is an older Dell PowerEdge 1600sc with four SCSI drives. Two are stand-alone, and the other two are in RAIDx (not sure the actual RAID id). It does not have an integrated RAID controller (nothing under integrated devices in the BIOS related to RAID), the client has a PowerEdge Expandable RAID controller PCI card for that. I can get into the RAID configuration by hitting <Ctrl><M> during the boot cycle, but this doesn't help me out too much since none of the drives are showing up (logical or physical). I just want to plug each individual SCSI drive to the motherboard's SCSI connector, then boot into DBAN and wipe them out.
I already got the two RAID'd drives scrubbed, but when I went to go do the second one DBAN returns a non-fatal error and will not proceed. I unplugged all the drives (which was my first mistake, since I can't remember what plugged in where) and then proceeded to attach just one at a time. However, the BIOS and boot report do not recognize that any drives are attached. I have the SCSI controller enabled under Integrated Devices in the BIOS, and the primary drive 0 & 1 set to auto, and they show 'Unknown Device' in the overview.
Here is a shot of the screen I am essentially stuck at:
I have limited SCSI drive experience, which is unfortunate for me, but hey, we've got to learn at some point or another, right?
Does anyone have any idea of how to hook the drive(s) up, set the BIOS, whatever it takes, to get the computer to recognize that I have the SCSI drive(s) attached?
Any and all help, tips, 'try this…' will be greatly appreciated. Thank you!
EDIT
Drive Models:
- MAP3735NP (Fujitsu)
- MXJ3073SC680600X (Worldisk (?))
- MBA3147NP (Fujitsu) (2 of these)
More pictures:
(note that there are 4 connectors total on this cable)
Best Answer
Taking a few minutes to go read the Wikipedia article on SCSI won't be a bad idea.
The "primary" and "secondary" are referring to the IDE controllers. You're not going to get SCSI disks to show up there.
It sounds like you've got fixed-mounted drives rather than hot-plug. If I'm wrong then this needs to be edited.
That machine has an integral SCSI controller on the motherboard. Its the LSI Logic controller that you see bannering its ROM messages directly below the errors about the primary and secondary IDE disks not being recognized in your screenshot. I'm fairly certain that DBAN will support the LSI Logic controller (w/ a Symbios Logic driver), whereas I'm not certain it will support the PERC. With that in mind, I'd ditch the PERC and use the motherboard controller.
Each disk should have a 68-pin female connection. There should be a cable with multiple male connectors to attach the disks to the controller. There is a 68-pin female connector on the motherboard for the integral SCSI controller (near the IDE connectors).
In a fixed-mounted configuration each disk has a jumper block to set the unique device ID. The disks should still be jumpered for unique IDs (so don't play w/ the jumpers). The last device connected to either end of a a SCSI bus must be a terminator. The motherboard SCSI controller will have an integral terminator. Not all hard disk drives can act as terminators, so your cable may have a terminator either integrated after the last female socket or attached to the last female socket. If the drives are capable of acting as a terminator then one of them will be jumpered for termination and should be connected last on the chain. (Post the drive model number in your question and we can link you up to the docs.)
Once you've got the drives connected to the motherboard controller you should see them detected by the controller.
Edit:
Perfect. You have an integral terminator on the cable. The disks should already be jumpered w/ unique IDs (don't move any jumpers now) since this was a working configuration.
With one disk attached to the SCSI cable and the cable attached to the motherboard SCSI connector (just as you showed in your photo) you should be able to see the disk detected by the LSI Logic SCSI controller during POST.
It's unclear to me if the disks were originally connected to the PERC or the LSI controller. I suppose it's possible, since the LSI is configured for ID 5 that one of the disks is also configured for ID 5 (the ID for each device, including the controller, must be unique on the SCSI bus), which would cause problems (typically showing no devices detected on the chain except the controller). If you find that one of the disks doesn't detect then, likely, you've got an ID conflict with that disk. You can always pull the ID jumpers completely, which will ID the disk as zero.