I replaced a failed SCSI tape drive on an OSR507 system and subsequently
get this message at start-up:

NOTICE: cha: SCSI bus has been reset ha=1
Attached SCSI peripherals will return to power up state \
(cha N04)
NOTICE: cha: Unexpected SCSI phase change \
ha=1 id=6 lun=0 cmd=12 blk=0 len=36
A SCSI peripheral behaved unexpectedly
Use Compaq diagnostics to confirm a hardware failure \
(chaN02)

After that, the new tape drive works fine!

The server is a 1999 Compaq Proliant 1600 with a dual channel integrated
Ultra-Wide SCSI-3 controller. The tape is the only device on the
secondary channel (non-RAID discs on the primary).

The old tape drive was a C1539 4/8GB DAT. This was connected to the SCSI
cable via an adapter since it was a plain SCSI-2 device with older style
connector (50 pin SE?). This had been working fine since 1999.

The new drive is a C5686B 40Gb DAT which is an Ultra Wide LVD device
with a 68-pin high density connector. The manual says it should work on
almost any SCSI bus (with performance degradation) except HVD. The
68-pin connector on the drive matched those on the existing SCSI cable
so I plugged it in directly. The SCSI cable has a terminator on the
end. Drive jumpers/DPI-switches were left at default (except SCSI ID
changed from 3 to 6 to match old drive).

The new tape drive was supplied with a SCSI cable which is clearly of a
later design (twisted pair ribbon in a tubular sleeve) than the 1999
proliant part (flat ribbon cable) but I didn't fit that since the
connector on the motherboard is relatively inaccessible.

Since the tape works fine in the tests I have done (3x faster than old
drive), I'm not sure whether I should worry about the error message.

Choices seem to be:
- Run HP diagnostics and see if anything is clarified.
- Replace the SCSI cable.
- ignore it and see what happens.

Has anyone encountered anything similar?
Can anyone explain what the error messages really mean?

Thanks for any thoughts on this.