[Info-vax] compressed SDLT capacity

B Hobbs bdhobbs18 at acm.org
Mon Nov 23 12:17:53 EST 2009


On Nov 23, 5:10 am, koeh... at eisner.nospam.encompasserve.org (Bob
Koehler) wrote:
> In article <36d33800-965f-4e70-a1c9-c2a309939... at u18g2000pro.googlegroups.com>, B Hobbs <bdhobb... at acm.org> writes:
>
> > I assume that when I do a INITIALIZE/ERASE, VMS sends a SCSI erase
> > command to the drive and VMS does other stuff until the drive signals
> > that the erase command is complete -- the drive does the work and the
> > SCSI bus is quiet.
>
>    INITIALIZE/ERASE is generic across a lot of things that don't
>    implement an erase command.  There's a very good chance that it's
>    writing a pattern to the tape.  If you use different options, I
>    think you can convince it to write more than one pattern.

With my VMS 7.1 treating the SDLT tape drive as a TK50, I suppose it
would depend on the SCSI commands that a TK50 drive could handle and
how the driver is programmed.  Anyone have a link to a TK50 manual
listing the SCSI commands?

>
> > How does a DIRECTORY of the SDLT work?  Does the tape drive send all
> > the data to VMS and VMS sorts thru the data to assemble something like
> > a directory listing?
>
>    DIRECTORY of a labeled tape reads data from the file headers.  On
>    a labeled tape there are three physical files for each logical file.
>    The first physical file contains the begining of file header.  The
>    second physical file contains the file data.  The third physical file
>    is the end of file header.
>
>    In order to get directory information from a tape, VMS reads the
>    first physical file, does a skip over the second file, and reads the
>    third physical file, for each logical file that shows up once in a
>    DIRECTORY listing.
>
>    A lot of modern tape drives are optimized for data flow, not
>    positioning.  Skipping files is a positioning operation.  There have
>    been drives supported by VMS that did not implement a skip operation
>    and VMS would actually read all the data in order to implement a
>    skip.

Again, I suspect I'm limited to what a TK50 can do.  Though my "TK50"
is faster and holds more than your TK50!  ;-)

Hmmm, I wonder if there is a way I can peek at what is happening on
the SCSI bus?  Ok, who took my tricorder?!?  It has ... err, unique
images of ... uhh, medical value of certain crew members.

-- Bill Hobbs



More information about the Info-vax mailing list