Finding rows count in a table without Select...

SQLServerCentral is supported by Red Gate Software Ltd.

Popular Topics

Home

Members

Calendar

Who's On

Home » Article Discussions » Article Discussions by Author » Discuss content posted by Manish Mittal » Finding rows count in a table without...

22 posts, Page 2 of 3

««1 2 3 »»

Finding rows count in a table without Select...

Rate Topic

Display Mode

Topic Options

Author

Message

garyg-755799

Posted Friday, March 26, 2010 10:05 AM

SSC Journeyman

Group: General Forum Members
Last Login: Monday, January 21, 2013 2:17 PM
Points: 98, Visits: 83

From my testing of these scripts, the row counts are only accurate if the table has a clustered index.

Gary

Post #890791

Greg Milner

Posted Friday, March 26, 2010 10:16 AM

SSC-Enthusiastic

Group: General Forum Members
Last Login: Thursday, May 16, 2013 11:36 PM
Points: 111, Visits: 410

I believe the WHERE clause here:
Where indid < 2
means that it's on the clustered index. So wouldn't it alway be accurate regardless of statics, since that's the actual physical layout of the rows?

Let me know if that's not correct.

Thanks.

G. Milner

Post #890809

Greg Milner

Posted Friday, March 26, 2010 10:21 AM

SSC-Enthusiastic

Group: General Forum Members
Last Login: Thursday, May 16, 2013 11:36 PM
Points: 111, Visits: 410

Hi, Manish.
I turned an email I wrote -- based on your post -- to a colleague (ETL designer) into a blog post on my site here, and cited you. Hope that's OK.

Thanks.

G. Milner

Post #890818

garyg-755799

Posted Friday, March 26, 2010 10:28 AM

SSC Journeyman

Group: General Forum Members
Last Login: Monday, January 21, 2013 2:17 PM
Points: 98, Visits: 83

Regarding the clause "Where indid < 2", if you have a clustered index there will be a row in sysindexes with indid = 1. If you do not have a clustered index, there will be a row in sysindexes = 0.

From my testing, only the tables with a clustered index (indid = 1) are accurate.

Gary

Post #890832

Jeff Moden

Posted Friday, March 26, 2010 9:16 PM

SSC-Dedicated

Group: General Forum Members
Last Login: Today @ 10:37 AM
Points: 32,893, Visits: 26,769

If you need more accurate results, try using DBCC UPDATEUSAGE first. If you haven't done it in awhile, it could take some good bit of time and will take less time the more you use it.

--Jeff Moden

"RBAR is pronounced "ree-bar" and is a "Modenism" for "Row-By-Agonizing-Row".

First step towards the paradigm shift of writing Set Based code:
Stop thinking about what you want to do to a row... think, instead, of what you want to do to a column."

For better, quicker answers on T-SQL questions, click on the following...
http://www.sqlservercentral.com/articles/Best+Practices/61537/

For better answers on performance questions, click on the following...
http://www.sqlservercentral.com/articles/SQLServerCentral/66909/

Post #891224

Chris Howarth-536003

Posted Saturday, March 27, 2010 1:24 AM

SSC Veteran

Group: General Forum Members
Last Login: Wednesday, May 08, 2013 3:01 AM
Points: 266, Visits: 1,023

DBCC UPDATEUSAGE - doesn't that only have to be run on SQL 2000 or upon upgrade from 2000 -> 2005?

I thought the issue had been corrected in 2005 onwards?

DBCC UPDATEUSAGE BOL quote:

"In SQL Server 2005, these values are always maintained correctly. Databases created on SQL Server 2005 should never experience incorrect counts, however, databases upgraded to SQL Server 2005 may contain invalid counts."

As I mentioned before, using sys.dm_db_partition_stats is the most reliable, future-proof method.

Chris

Post #891263

Jeff Moden

Posted Saturday, March 27, 2010 12:58 PM

SSC-Dedicated

Group: General Forum Members
Last Login: Today @ 10:37 AM
Points: 32,893, Visits: 26,769

Chris Howarth-536003 (3/27/2010)

Actually, that's exactly correct, Chris. I've been stuck in a 2k world for much too long. Being in a bit of a hurry, all I saw was a reference to sysindexes and the word "inaccurate" and didn't realize folks were talking of 2k5+.

--Jeff Moden

Post #891326

Johnc

Posted Monday, March 29, 2010 2:11 AM

Grasshopper

Group: General Forum Members
Last Login: Tuesday, December 07, 2010 2:34 AM
Points: 13, Visits: 31

DBCC UPDATEUSAGE - doesn't that only have to be run on SQL 2000 or upon upgrade from 2000 -> 2005?

Actually that may provide the explanation. My original post was about inaccurate counts in a 2000 database I had migrated to 2008 and I didn't know about the caveat, so that may well be the answer.

Thanks Jeff, Chris (and All)

Post #891641

Yanko Nikolov

Posted Wednesday, March 31, 2010 12:42 AM

SSC-Enthusiastic

Group: General Forum Members
Last Login: Tuesday, April 30, 2013 1:50 AM
Points: 100, Visits: 148

Hello
This is the cool example, but you do not need to write own procedure to determine the tables row count. Just try this:

exec SP_MSFOREACHTABLE @command1=N' exec SP_SPACEUSED ''?'' '

Regards
Yankov Yanko

Post #893502

krassimir.kotzev

Posted Friday, April 09, 2010 12:35 PM

Forum Newbie

Group: General Forum Members
Last Login: Friday, April 30, 2010 11:50 AM
Points: 1, Visits: 24

And th equestion again: How recent are your statistics, indexes, etc. ? If you can not tell, better use count(*) to get the numbers. It is slow but accurate.

Post #900766

« Prev Topic | Next Topic »

22 posts, Page 2 of 3

««1 2 3 »»

Permissions