This message was intended for James Overton at something
like HP.ROSE.COM, but the reply bounced, so ....

Hello James,

> : But my original question remains--why does this happen
> : when there has been fairly frequent use in compress mode
> : over several days (weeks), no human intervention, and
> : no intervening reboot?
> <snip>
> Our experience has been that a power glitch to the tape drive (external
> units) will cause the reset to uncompressed.  It is a good idea to
> Always do a DEVCTRL to set the device to the desired configuration before
> writing to the DDS.  (an ounce of prevention....).

This is an embedded drive.  Surely a power glitch would have caused
the backup to stop dead in its tracks, if not halt the processor
as well.  I don't yet have the UPS that came with the machine
(an HP3000/928) connected--that will happen tomorrow.

We do not use Oracle on our HP3000s; some people have suggested
that Oracle DBs do not have much compressible data, but in any
event that can't be the problem.

We have two identical drives on our production machine (a 957)
and have never experienced this problem with those drives.

I'll do some experimenting when time permits.  There are a number
of things to explore.  For a start, the media in use at the time
of the silent switch to uncompressed mode is a fairly old Verbatim
60-metre tape.  Maybe it doesn't perfectly match DDS media specs.

Surely it couldn't have anything to do with the fact that this
machine began life as an HPUX box?  Could it still be harbouring
fragments of that virus?

I promise to report back to the list if anything clear and
reproducible turns up.

Cheers, and many thanks for your comments.

Ron Burnett
[log in to unmask]