We use 'spans' and remove them in one operation and also do not commmit
anything until we finish a pass over all rows.

But main trick is blocked views, which uses smaller footprint on commits.
Murat

-----Original Message-----
From: Brian Kelley [mailto:[EMAIL PROTECTED]
Sent: Wednesday, May 12, 2004 1:55 PM
To: Metakit mailing list
Subject: [Metakit] First metakit failure, database grew to 2+ gigabytes


This is just a FYI: I had my first metakit failure today that
corrupted the database.  I have a process that deletes objects from
tables by finding the object id, searching for it in the table and
then deleting it wholesale.

It wasn't intelligent enough to find table "spans" to remove blocks of
objects at a time, so it was deleting a row at a time and commiting
the data occasionally.  The database started at 363MB and grew to
2,097,153KB before the views got corrupted.  Before that, it was as
happy as a clam and chugging away.

So I guess that I now know the size limit of mk databases.  The take
home message is deleting rows can rapidly grow memory.  Is there an
alternate strategy that I can use in this case for deleting rows from
a database?

Anyway, I can keep better track of database size know that I know the
limitations a little better.

Brian

_____________________________________________
Metakit mailing list  -  [EMAIL PROTECTED]
http://www.equi4.com/mailman/listinfo/metakit
_____________________________________________
Metakit mailing list  -  [EMAIL PROTECTED]
http://www.equi4.com/mailman/listinfo/metakit

Reply via email to