[weblacky]

Yeah, I agree. If the original platform confirms it’s borked then, no use going other routes.

[Raion]

What I meant was assuming the rest of the system was in good working order - clearly it's not working to grab the drive this way and unless he wants to pay a data recovery service thousands to recover the data, he can be assured that the litmus will be powering it up.

Obviously, if you value your 4D era systems, don't power them up without going in to fix the stuff that's known to degrade on them.

[jan-jaap]

NB: If you open the chassis of the 4D twin tower, you will learn about the "intrusion detection" switches. Basically, if you open the access panel of the main tower or remove the disk stack from the small tower and the system is powered up, the breaker switch will cut the power. Especially the main tower panel can be an annoyance, but it's not too hard to work around.

Finally: "This Old SGI" mentions that PSU problems can manifest themselves as disk problems! But if your disk doesn't function when attached to another system that's obviously not the problem.

[iansmith]

I've had the main tower open before (and opened it just not to check the condition) and have pulled the boards 20 years ago when it booted up and was functional.

So now I have to decide how much work I want to put into this to see if it can power up again. I mostly want to archive that hard drive because who knows what might be on it and I'd like to make it available to others to use as an image. Plus I just kinda want to see what's on it.

I don't want to blow the thing up trying to power it on though. I wouldn't mind selling it to someone qualified to work on them, it's not doing anyone any good sitting in my garage. Are they still in demand? I have no idea how I would ship this thing.

[Raion]

Private message CB_HK. He may be able to help you!

[miod]

Linux reports...

[sda] Read Capacity (1) failed: Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[sad] Sense Key : Not Ready [current]
[sda] Add. Sense: Logical unit not ready, cause not reportable

Is this an issue with my setup, or does this look like the drive has failed?

In order to properly identify an SCSI device, the operating system must send a TUR (Test Unit Ready) command to the device. In case of a hard disk, this will cause the motor to start if it wasn't started, and report its parameters. If the operation fails (for example because the disk has not been able to spin up correctly), no parameters gets returned and the disk can not be used.

With such an old disk, it is very likely that the spindle can not reach its nominal speed fast enough. You might want to check that the power applied to the disk is correct; also you might want to gently rock the disk (while not powered on) in case something is stuck inside.

[iansmith]

In order to properly identify an SCSI device, the operating system must send a TUR (Test Unit Ready) command to the device. In case of a hard disk, this will cause the motor to start if it wasn't started, and report its parameters. If the operation fails (for example because the disk has not been able to spin up correctly), no parameters gets returned and the disk can not be used.

With such an old disk, it is very likely that the spindle can not reach its nominal speed fast enough. You might want to check that the power applied to the disk is correct; also you might want to gently rock the disk (while not powered on) in case something is stuck inside.

I can hear the disk spinning when I power it up, but no head movement. I can't tell if it's spinning at full speed or not, but the hum is very even. I checked the voltages while it's running and they are right at 5v and 12v and the amp draw matches the idle current the specs list. I think I am giving it enough power but can't know for sure.

I've used the sg3-util package as suggested to try and get the drive to fully spin up. I just tried sg_turs and it reports the same information immediately. Drive makes no noise. I don't see any power fluctuations when I send those commands.

Would termination be an issue? The drive itself has no terminators, I've tried it with and without a terminated device after the drive with the same results. From what I recall, bad termination means drives simply don't show up. I can mount other SCSI devices on the chain so I know communication is working.

I remember when Seagate had problems with their drives, I had to whack a few on the side when they were trying to spin up to get them going but not going to do that to this one, heh. The platters are clearly spinning so the heads aren't stuck anywhere.

I have a SCSI2SD device I can use to boot the Iris if it's still operational and I assume I can download installation media somewhere. Still, I'd like to get the drive copied so I'll continue exploring options.

[jan-jaap]

I have no idea how I would ship this thing.

Strap it on a pallet, crate it and use freight shipping. The problem is cost (and effort). Usually it's a much better idea if someone local comes and picks it up.

If there really are no takers I'd be interested in some of the bits to keep my own system going, but if someone wants the whole systems that would be a much better idea.

[miod]

I can hear the disk spinning when I power it up, but no head movement. I can't tell if it's spinning at full speed or not, but the hum is very even.  I checked the voltages while it's running and they are right at 5v and 12v and the amp draw matches the idle current the specs list.  I think I am giving it enough power but can't know for sure.

I've used the sg3-util package as suggested to try and get the drive to fully spin up. I just tried sg_turs and it reports the same information immediately. Drive makes no noise. I don't see any power fluctuations when I send those commands.

Then I'd suspect the on-disk electronics (the controller board).

Would termination be an issue? The drive itself has no terminators, I've tried it with and without a terminated device after the drive with the same results. From what I recall, bad termination means drives simply don't show up. I can mount other SCSI devices on the chain so I know communication is working.

Since the disk gets recognized with a unique ID and answers commands, there is no cabling or termination problem.