Solaris LUNs

[popo]

Howdy all,

I have just added a new scsi disk to
a SPARC5 box.

From the PROM when a do a:

probe-scsi-all

Target 0 (the target which this disk is on)
is detected but along with it is listed
7 LUNs for the *same* disk.

Has anyone had any experience with probe-scsi-all detecting multiple LUNs for
a "hard-disk" and why the PROM would detect more than 1 LUN for a HD ?

Cheers

- Alex

 

[Nothinman]

Are you sure the termination is done properly?

 

[popo]

Seems to be all terminated correctly. ??

 

[LNXman]

Little desc:

probe-scsi: This will test all devices connected to the on board SCSI controler.
probe-scsi-all: This will test all devices connected to all the SCSI controlers.

I think you want to use probe-scsi and not probe-scsi-all.

Just a thought.

GL

 

[Nothinman]

I would then check the id jumper.

 

[popo]

Same deal for both probe-scsi and probe-scsi-all

 

[popo]

<< I would then check the id jumper. >>




There are no manual jumpers set. ie default.
I have tryed setting the jumpers manually....same deal though.

 

[LNXman]

<<
There are no manual jumpers set. ie default.
I have tryed setting the jumpers manually....same deal though. >>



If you have no jumpers set, then you are using ID 0, which mostlikely is being used by your boot drive (if you already had a drive), thus causing a conflict.
You should make sure you are not conflicting with another ID as well as having the correct termination.
i.e. An on-board device chain will ONLY have the last scsi device terminated. External devices will ONLY have the last external scsi device terminated in the chaing as well. Unless the on-board chain is continuous with external device chain in which case the last external device will be the only one terminated.

You could verify if your machine detects anything at all when you boot it (boot -r). It should give you a hint of what is going on.

GL

/edit: fixed typo

 

[popo]

<<

<<
There are no manual jumpers set. ie default.
I have tryed setting the jumpers manually....same deal though. >>



If you have no jumpers set, then you are using ID 0, which mostlikely is being used by your boot drive (if you already had a drive), thus causing a conflict.
You should make sure you are not conflicting with another ID as well as having the correct termination.
i.e. An on-board device chain will ONLY have the last scsi device terminated. External devices will ONLY have the last external scsi device terminated in the chaing as well. Unless the on-board chain is continuous with external device chain in which case the last external device will be the only one terminated.

You could verify if your machine detects anything at all when you boot it (boot -r). It should give you a hint of what is going on.

GL

/edit: fixed typo >>




There is only one HD and yes it is set ID 0. There are no internal HDs.
There are no conflicts.

 

[LNXman]

Do you have a CDrom drive? Is it external? Internal? If the cdrom is internal, then you should not have to terminate the drive (if you have done so). I wish I had a solaris machine with me to re-create your problem.
I am really running out of options.

The last thing I could suggest if there are no conflicts, or probs, is to just to have the machine re-scan all scsi devices by rebooting it with boot -r at the bios prompt, stopping it again, and then re-doing a probe-scsi again.

GL

 

[DirtylilTechBoy]

What do you mean by Termination???

 

[Nothinman]

SCSI devices need the chain to be terminated because you can have an arbitrary number of devices (depending on the card up to 15, counting the controller as 1).

www.scsifaq.com has a lot of SCSI information.