It is/was a timing issue where the tape is being passed back to the
DYNAMCMS server at DYNCLOSE to write trailer labels for the dataset ...
then back to the end user who subsequently issues a DYNOPEN for DSN 4 on
the same tape causing the tape to be passed back to the DYNAMCMS server
to validate the volume ... rewind to check VOL1 followed by subsequent
FSF(s) to position to the end of DSN 3, then prepare the HDR1 labels for
the new 4th DSN.

In any case, the timing issue caused DYNAMCMS to incorrectly state there
was only one file on the tape resulting in a automatic CANCEL ... so the
tape was never reattached to the end user because the DYNOPEN was
cancelled.  This resulted in DDR failing because there actually was no
tape attached to the end user at virtual device 181.

I've been waiting for another report of a failure from the customer
after providing them with some trace diagnostics ... but as far as I
know there hasn't been another failure (because I haven't heard anything
back for over a week).

JR (Steven) Imler
CA
Senior Software Engineer
Tel:  +1 703 708 3479
Fax:  +1 703 708 3267
[EMAIL PROTECTED]


> -----Original Message-----
> From: The IBM z/VM Operating System 
> [mailto:[EMAIL PROTECTED] On Behalf Of Alan Altmark
> Sent: Friday, May 02, 2008 01:22 PM
> To: [email protected]
> Subject: Re: Problem with stacked ddr tape and possible DYNAM
> 
> On Friday, 05/02/2008 at 12:06 EDT, Jan Canavan 
> <[EMAIL PROTECTED]> 
> wrote:
> > What we get on the previous volumes are:
> > 
> > --------------------------------------------
> > Tape 0590 given to DYNAMVM 0590
> > Tape 0181 attached
> > 11:22:13     CADT831I *CLOSED* DDRWKLY 0590 050094 DDR.XX.530W01.2
> > 
> ----------------------------------------------------------------------
> > 
> > You don?t see the TAPE 181 ATTACHED.
> > 
> > If you add sleep of 180 seconds, or you put a trace in it 
> works.  DYNAM 
> support 
> > cannot reproduce the problem.
> > 
> > Anybody else had trouble with this?
> 
> Perhaps DYNAMVM is look for the "given" message instead of 
> waiting for the 
> interrupt that indicates a device has appeared?  (I don't know how it 
> works.)
> 
> Alan Altmark
> z/VM Development
> IBM Endicott
> 
> 

Reply via email to