RE: index doesn't contain duplicates

From: Goulet, Richard <Richard.Goulet_at_parexel.com>
Date: Tue, 9 Jun 2009 14:22:13 -0400
Message-ID: <23C4836D8E9C5F4280A66C0C247BC16F2972C15A_at_US-BOS-MX011.na.pxl.int>

Josh,  

    One question & this one has taken me to the cleaners more times that I'd like to admit. The two rows that "appear" to be the same, are they of the same length? Here's an example of what I mean:  

    Row1: 'DATAVALUE'
    Row2: 'DATAVALUE '  

There is a very slight difference, in this case a trailing space that really does make the two rows different. Other characters that can cause similar problems is anything in the extended ascii character set which will show up like spaces.  

Dick Goulet
Senior Oracle DBA
PAREXEL International  

---

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Josh Collier
Sent: Tuesday, June 09, 2009 2:12 PM
To: mark.powell_at_eds.com; oracle-l_at_freelists.org Subject: RE: index doesn't contain duplicates

Version 9208  

The way the table was created was this..  

Table was created with a CTAS

Non-unique index was built

Primary key was built on a column, specifying index created in step2. Enable novalidate. The data in question was inserted after the primary was in place.      

The plan that produces the incorrect results is a fast full scan on the non-unique index

The plan that produces the correct result is a full tablescan on the table.  

We  

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Powell, Mark D
Sent: Tuesday, June 09, 2009 10:22 AM
To: oracle-l_at_freelists.org
Subject: RE: index doesn't contain duplicates  

Yes, the full Oracle version would be nice plus the select with plan the produces the proper results and one that does not.  

Things you can try:  

Run analyze index validate

Run analyze table validate cascade (depending on results from above)

Drop and re-create index  

• Mark D Powell -- Phone (313) 592-5148

  ---

        From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Josh Collier

	Sent: Tuesday, June 09, 2009 12:11 PM
	To: oracle-l_at_freelists.org
	Subject: index doesn't contain duplicates

	All,

	 

	I have an table that has duplicate values for a column. I have a

non-unique index on this column. When I run a query that uses a fast full scan to retrieve the rows, it only retrieves one of the two rows. When I force a full tablescan both of the rows come back. I have used all my sql tricks to try my best to determine if the value has trailining or leading non-printable characters and it doesn't appear so. I have also moved the values into a varchar 11 field in another table, the value is 11 chars long. Both of the duplicate rows where accepted. Does anyone know of a bug or a reason why both of the rows wouldn't appear in the index?          

        Thanks for your time,

        Josh C.

--
http://www.freelists.org/webpage/oracle-l

Received on Tue Jun 09 2009 - 13:22:13 CDT