RE: 8.1.7.2.5 interMedia Text Slow Sync - SOLVED

2002-01-31 Thread Jack C. Applewhite

For what seems to be the few of you who use interMedia Text, I found the
problem - it was ME. :-(

I had set (in CTX_Parameters) Max_Index_Memory to 200MB (from the default of
12.5MB) and used 200MB when I created indexes, but had overlooked
Default_Index_Memory (default of 12.5MB), which is used by
CTX_DDL.Sync_Index.

After I set Default_Index_Memory to 200MB, re-syncs jumped from about 750
CLOBs per minute to over 2,500 per minute.  Oracle was able to sort an
entire batch of 157,000 CLOBs in memory, instead of only about 8,000 at a
time - surprise, surprise with 16 times the memory available.

BTW, it didn't hurt that I spread the datafiles of the tablespaces that hold
the DR$$X segments across 2 drives, instead of just 1.  That boosted
re-syncs from 600 CLOBs per minute to 750.  I/O distribution never hurts.

Jack


Jack C. Applewhite
Database Administrator/Developer
OCP Oracle8 DBA
iNetProfit, Inc.
Austin, Texas
www.iNetProfit.com
[EMAIL PROTECTED]
(512)327-9068


-Original Message-
Applewhite
Sent: Thursday, January 24, 2002 9:35 AM
To: Multiple recipients of list ORACLE-L


We are experiencing very slow interMedia Text resyncs ( CTX_DDL.Sync_Index )
under 8.1.7.2.5 on Win2k Server.

We upgraded from 8.1.6.0.0 to 8.1.7.2.5 earlier this month (both under Win2k
Server) on essentially the same hardware.  We used to see about 700-2000
CLOB documents per second indexed under 8.1.6, but now see only about
400-800.  The low-high numbers relate to the size of the index being
sync'd - low just before we roll out a monthly partition (total 2 million
CLOBs), high just after (total 1 million CLOBs).

Query performance is better under 8.1.7 - partly because I spread the
DR$$I table across 3 drives.  By the looks of the I/O pattern, I should
spread DR$$X (the index on DR$$I table) across 2 or 3 drives as well.

Anybody have any words of wisdom as to what I can do to speed things up?

BTW, I'm curious at how 8.1.7 balances I/O.  While it's writing to the
DR$$X segment, it doesn't write to the online redo logs - it waits until
it's reading from DR$$X segment before it writes to the redo logs.  I
wonder why it doesn't do the writes to redo in parallel with the writes to
DR$$X.

Thanks.

Jack


Jack C. Applewhite
Database Administrator/Developer
OCP Oracle8 DBA
iNetProfit, Inc.
Austin, Texas
www.iNetProfit.com
[EMAIL PROTECTED]
(512)327-9068




-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.com
-- 
Author: Jack C. Applewhite
  INET: [EMAIL PROTECTED]

Fat City Network Services-- (858) 538-5051  FAX: (858) 538-5051
San Diego, California-- Public Internet access / Mailing Lists

To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).



8.1.7.2.5 interMedia Text Slow Sync

2002-01-24 Thread Jack C. Applewhite

We are experiencing very slow interMedia Text resyncs ( CTX_DDL.Sync_Index )
under 8.1.7.2.5 on Win2k Server.

We upgraded from 8.1.6.0.0 to 8.1.7.2.5 earlier this month (both under Win2k
Server) on essentially the same hardware.  We used to see about 700-2000
CLOB documents per second indexed under 8.1.6, but now see only about
400-800.  The low-high numbers relate to the size of the index being
sync'd - low just before we roll out a monthly partition (total 2 million
CLOBs), high just after (total 1 million CLOBs).

Query performance is better under 8.1.7 - partly because I spread the
DR$$I table across 3 drives.  By the looks of the I/O pattern, I should
spread DR$$X (the index on DR$$I table) across 2 or 3 drives as well.

Anybody have any words of wisdom as to what I can do to speed things up?

BTW, I'm curious at how 8.1.7 balances I/O.  While it's writing to the
DR$$X segment, it doesn't write to the online redo logs - it waits until
it's reading from DR$$X segment before it writes to the redo logs.  I
wonder why it doesn't do the writes to redo in parallel with the writes to
DR$$X.

Thanks.

Jack


Jack C. Applewhite
Database Administrator/Developer
OCP Oracle8 DBA
iNetProfit, Inc.
Austin, Texas
www.iNetProfit.com
[EMAIL PROTECTED]
(512)327-9068



-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.com
-- 
Author: Jack C. Applewhite
  INET: [EMAIL PROTECTED]

Fat City Network Services-- (858) 538-5051  FAX: (858) 538-5051
San Diego, California-- Public Internet access / Mailing Lists

To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).