Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!dev!dgis!jkrueger
From: jkrueger@dgis.dtic.dla.mil (Jon)
Newsgroups: comp.databases
Subject: Re: Help with RTIngres 5.1
Keywords: quel, ingres, help
Message-ID: <699@dgis.dtic.dla.mil>
Date: 14 Dec 89 00:24:47 GMT
References: <590@aut.UUCP> <1989Dec13.231419.16400@athena.mit.edu>
Organization: Defense Technical Information Center (DTIC), Alexandria VA
Lines: 33

joe@athena.mit.edu (Joseph C Wang) writes:

>I have a table of about 40,000 rows.  Whenever I try to add about
>10,000 more rows via the copy command the database freezes up for
>hours at a time if the table is organize as a chash or cbtree table.
>I get around this be modifying the table to cheap, adding the rows,
>and modifying it back to chash or cbtree.  Is there a better (i.e.
>faster) way?

No, you're doing exactly the right thing.  Otherwise INGRES has to
maintain the index structure on the table during the copy operation.

From the INGRES/QUEL Reference Manual (Release 6, UNIX, 1989), Section
2.5.2.1: "Copying from a file into a non-journaled heap table without
secondary indices runs significantly faster than copying into a btree,
isam, or hash table, or one that is journaled or has secondary
indices ... You may get better performance by modifying [the table] to
heap before doing the copy and then modifying to the correct structure
when the copy is complete."

>Also are there books on optimizing Ingres, the technical notes seem
>a bit sketchy.

The DBA Guide and the course notes from the Performance class
are good.  Talks by Ingres Corp. people at user group meetings
also: "Top 10 Performance Mistakes", "Top 10 Concurrent Performance
Mistakes".  These are reprinted in IUA newsletters.

-- Jon
-- 
Jonathan Krueger    jkrueger@dtic.dla.mil   uunet!dgis!jkrueger
The Philip Morris Companies, Inc: without question the strongest
and best argument for an anti-flag-waving amendment.