ADSM is not deciding what files to delete, as this is a function of the
DataTools SQL-Backtrack software.  Within SQL-Backtrack configuration,
you determine how many copies of the database there will be.

Mark Mapes
PG&E

----------------------[Reply - Original Message]----------------------
Sent by:Tom Denier <tom AT WAL6000B.UDC.UPENN DOT EDU>
Sent by:Tom Denier <tom AT WAL6000B.UDC.UPENN DOT EDU>
 Cyndie Behrens wrote:
> We are currently updating the SQL-BackTrack section of the ADSM
redbook,
> SG24-4335, and this is one area we want to document.  When an api
application
> uses a different name each time it sends an object to ADSM, that means
that
> the object stored on the ADSM server is always active, and hence,
never
> expires.

If this were really true the amount of data in our API filespaces would
grow linearly with time. This is not happening; the amounts of data are
stable most of the time, and go up a bit when the database managers
add or enlarge databases. There is clearly some mechanism causing
ADSM
to decide that the API files have been deleted from the host system
and eventually drop them from the storage pools under control of the
RETONLY parameter.

============================================================
=========
=========