Hi Gary,
 
The first, really easy time-saver is.
turn off your image-logging before you start deleting.
(having a backup is a good idea)
This really helps.
 
<PLUG ALERT>
Robelle consulting sell SUPRTOOL.
I have to say they have not done anything exiting in the past 3 years.
(except supplying us with stories about their summer-houses on tropical islands)
However suprtool is really fast in extracting data from databases, and will put
back in another database very very very fast. If the selection criteria arent too complex I think this is a very good solution.
Robelle will probably be thrilled to send you a demo tape.
</PLUG ALERT>
 
Cheers,
 
Leon Degeling
Amsterdam
 
 
 
----------
From:   Gary Clarke[SMTP:[log in to unmask]]
Sent:   donderdag, januari 11, 1996 5:12
To:     Multiple recipients of list HP3000-L
Subject:        HP3000 Image: Mass Data Delete
 
I have a need to delete large amounts of data from some databases
(largest db is >30 million sectors) according to certain simple logic
algorithms. This is to be done on a regular basis in order to
down-size our production databases for our development environment.
 
Currently, I have some C programs deleting the data. A couple of them
take more than 10 days to run.
 
Can anybody suggest a more efficient method for deleting this data,
e.g. tools that may be available.
 
Approximately 3/4th of the total data is deleted so I was considering
extracting what I wanted, purging the appropriate sets, rebuilding
them as empty with Adager, and then importing the data. However, I'm
not sure if the time savings would justify the effort.
 
The jobs are generally waiting on disk i/o accoring to our technical
support, cpu cycles is not a problem.
 
Any ideas would be appreciated.
 
Thanks, Gary Clarke
 
[log in to unmask]