[Previous entry: "Post about setting up and using autotrace"] [Next entry: "Sitemap generation tweaked"]
Disabling Oracle writes into NT event log
December 20th, 2004 by Pete
Post to del.icio.us
Post to Furl
I found an interesting post on the Oracle-l list last week about disabling writes to the Windows event log. Because I like undocumented or hidden or hard to find data I was immediately attracted to this post. The thread is called "disabling Oracle writes into NT event log" and the original poster said every time an archivelog is created an event is written to the event log in the same manner as an entry is written to the alert log.
The poster wanted to disable this write to the event log and found an undocumented parameter _disable_ntlog_events. You can test the default value of this parameter with my script check_parameter.sql. An example for my 9iR2 personal edition database is given next:
check_parameter: Release 1.0.0.0.0 - Production on Mon Dec 20 15:27:29 2004
Copyright (c) 2004 PeteFinnigan.com Limited. All rights reserved.
PARAMETER TO CHECK [utl_file_dir]: _disable_ntlog_events
CORRECT VALUE [null]: FALSE
OUTPUT METHOD Screen/File [S]: S
FILE NAME FOR OUTPUT [priv.lst]:
OUTPUT DIRECTORY [DIRECTORY or file (/tmp)]:
Investigating parameter => _disable_ntlog_events
====================================================================
Name : _disable_ntlog_events
Value : FALSE
Type : BOOLEAN
Is Default : DEFAULT VALUE
Is Session modifiable : FALSE
Is System modifiable : FALSE
Is Modified : FALSE
Is Adjusted : FALSE
Description : Disable logging to NT event log
Update Comment :
-------------------------------------------------------------------------
value is correct
PL/SQL procedure successfully completed.
For updates please visit http://www.petefinnigan.com/tools.htm
SQL>
The poster changed the value of this parameter with an ALTER SYSTEM command and checked it had been changed. He then noted that the events were still being written. His version was 10.1.0.2.0 on Windows XP. Niall Litchfield answered with information that he thought this bug was fixed in 10.1.0.3.
As I said at the start of this entry I found this interesting because it talks about internals, in this case an undocumented or underscore parameter.
