U320 Drives Show up on Wrong Target ID?
Paul Berger
phb.hfx at gmail.com
Sun Jun 5 09:28:18 CDT 2016
On 2016-06-05 6:25 AM, Jonathan Katz wrote:
> On Sun, Jun 5, 2016 at 7:26 AM, <alexmcwhirter at triadic.us> wrote:
>> So in short i have six identical drives. Same manufacture, model, and
>> firmware. Three of them work fine, the other three also work fine but they
>> always show up as the wrong target.
> Is there a jumper on the drives that is overriding the address
> assigned by the enclosure? The whole point of SCA is that you don't
> use jumpers to assign the drive address, it just figures it out
> somehow. It sounds like the SCA isn't being allowed to do its thing.
The SCA connectors has pins on it for setting the drive address, so the
the disk backplane slot that the drive is plugged into sets the drive
address. These drives normally have jumper positions on them to set
addresses as well, however the normal case is that the actual address is
the backplane set address ORed with what is set in the jumpers. For
instance if the drive has a jumper for address 2 set, and you plug it
into the address 0 slot it will respond to address 2, but if you plug it
into the address 1 slot it will respond to address 3, but when plugged
into slot with address 2 it will correctly respond to 2. The OP claims
that these
drives respond to the same address no matter where they are plugged,
which does not seem possible on any SCA drive I have ever seen.
If the drives happen to have 50 or 68 pin connectors that are mounted on
a carrier with a connector that plugs into a backplane that sets an
address, in this case this is normally done with a small cable that
connects the auxiliary connector on the drive to the appropriate pins on
the carrier, in that case if this cable is disconnected and address
jumpers are installed on the drive, I could see the OP's situation
occurring.
I have no experience with Sun disk enclosures so I don't know what sort
of drives or drive carriers they use in their enclosures so I cannot say
for sure what may be going wrong.
Paul.
More information about the cctech
mailing list