Pete Turnbull wrote:
On 11/05/2007 10:06, Jules Richardson wrote:
A 31 hour restore if there's 2.3GB of data on
the tape is a bit on the
lengthy side...
Indeed, I'm sure mine used to take about 2-3 hours. It couldn't be
something to do with synchronous vs async, could it?
I'm not sure. It's logged as connected with a user speed of 10MB/s, but
"goal"
and "current" speeds of 3.3MB/s. But still, 3.3MB/s is a long way off 21KB/s -
even allowing for SCSI protocol overheads :-) It certainly didn't say at boot
time that it was async, anyway.
The TTi adapter board inside the drive seems to be an interceptor mainly for
the front panel - SCSI in from the HBA on the one side, then SCSI out to the
Exabyte EXB-8500 drive on the other. In other words, I don't know of a way of
seeing what the bus between the intercept board and the drive itself is doing
- it's quite possible there's a problem there.
Unfortunately the EXB-8500 drive seems to have been modified firmware-wise as
part of the TTi setup; it refuses to talk direct to the HBA and doesn't even
show up in an inquiry at boot time. Looks like it'll *only* cooperate with the
TTi intercept board. Either that or it's utterly, utterly broken.
The only other possibilities are the unknown jumpers on the back of the main
unit are running the drive in some sort of restricted mode, or there's some
weird vendor-unique command that needs to be sent to the drive to kick it into
high gear.
If someone has a known-working one of these TTi units and can at least quote
me their jumper settings off the back, that would be useful! 10-12 control the
SCSI ID (as someone's scribbled that on the drive in pen), but I have no idea
what the other 9 are for.
If it comes to it, I'll just leave the thing running for a day and a half...