Delete duplicate records in a table

Question

Post reply

Delete duplicate records in a table

chitturiii

SSC Enthusiast

Points: 185
More actions
December 1, 2010 at 2:33 am

#243410

Hi,
Can you please suggest me the best query to delete the duplicate records in a table with large number of records.
Thanks in advance.
--Chandra

Viewing 9 posts - 1 through 8 (of 8 total)

You must be logged in to reply to this topic. Login to reply

VSSGeorge SSCrazy Eights Points: 8151 More actions · Answer 1

Hi ..

Try this one...

declare @tab table (id int,name varchar(10))

insert into @tab (id,name)

select 1,'Sumit' union all

select 2,'Sumit2' union all

select 2,'Sumit2'

select id,name from ( select row_number() over (partition by id,name order by id )as row ,id,name from @tab )as t

where row= 1

ChrisM@Work SSC Guru Points: 186127 More actions · Answer 2

chitturiii (12/1/2010)
Hi,
Can you please suggest me the best query to delete the duplicate records in a table with large number of records.
Thanks in advance.
--Chandra

You may have to delete in batches if there are a very large number of rows to be deleted. Do you have a query which identifies the rows to delete? If not, post the structure of the table as CREATE TABLE etc and any indexes, also which columns are used to identify dupes.

^{“Write the query the simplest way. If through testing it becomes clear that the performance is inadequate, consider alternative query forms.” - Gail Shaw}

For fast, accurate and documented assistance in answering your questions, please read this article.
Understanding and using APPLY, (I) and (II) Paul White
Hidden RBAR: Triangular Joins / The "Numbers" or "Tally" Table: What it is and how it replaces a loop Jeff Moden

PKDC SSChasing Mays Points: 618 More actions · Answer 3

If you want to filter duplicate, just use group by clause to filter duplicate row.

select column_name, min(PK_Column_name) PrimaryKeyColumn from tablename group by column_name having count(*)>1

The above script will identify duplicate value and show minimum Primary key value. If you intend to keep one row, you can keep the minimum Primary key value row. (of course you can change the condition)

The below will delete all duplicate rows and leaving one distinct value row(minimum Primary key value row)

with Tempinfo (columename, PrimaryKeyColumn)

as

(select column_name, min(PK_Column_name) PrimaryKeyColumn from tablename group by column_name having count(*)>1)

delete a

from tablename a join Tempinfo b on a.column_name=b.column_name and a.PrimaryKeyColumn<>b.PrimaryKeyColumn

Sachin Nandanwar SSCertifiable Points: 7699 More actions · Answer 4

Also it is a better idea to put the recover model to SIMPLE or else your log file will bloat.But make sure you take a Full and a T log backup before changing the recovery model to SIMPLE.

--------------------------------------------------------------------------------------------------
I am just an another naive wannabe DBA trying to learn SQL Server

Jeff Moden SSC Guru Points: 1004284 More actions · Answer 5

Sachin Nandanwar (12/3/2010)
Also it is a better idea to put the recover model to SIMPLE or else your log file will bloat.But make sure you take a Full and a T log backup before changing the recovery model to SIMPLE.

Agh... be careful now. Changing to SIMPLE breaks the backup chain. This is normally one of the worst things you can do.

--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.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)

Sachin Nandanwar SSCertifiable Points: 7699 More actions · Answer 6

Jeff Moden (12/4/2010)
Sachin Nandanwar (12/3/2010)
Also it is a better idea to put the recover model to SIMPLE or else your log file will bloat.But make sure you take a Full and a T log backup before changing the recovery model to SIMPLE.
Agh... be careful now. Changing to SIMPLE breaks the backup chain. This is normally one of the worst things you can do.

But that's the reason I suggested to take a FULL backup before changing it to SIMPLE.Then delete the records.Change it to FULL.Again take a Full Backup.

Correct me if I am missing something.

--------------------------------------------------------------------------------------------------
I am just an another naive wannabe DBA trying to learn SQL Server

Hardy21 SSCrazy Eights Points: 9708 More actions · Answer 7

Sachin Nandanwar (12/7/2010)
Jeff Moden (12/4/2010)
Sachin Nandanwar (12/3/2010)
Also it is a better idea to put the recover model to SIMPLE or else your log file will bloat.But make sure you take a Full and a T log backup before changing the recovery model to SIMPLE.
Agh... be careful now. Changing to SIMPLE breaks the backup chain. This is normally one of the worst things you can do.
But that's the reason I suggested to take a FULL backup before changing it to SIMPLE.Then delete the records.Change it to FULL.Again take a Full Backup.
Correct me if I am missing something.

Execute the query inside the Begin transaction - End Transaction, If you satisfy with the result then commit it otherwise rollback it. So, it will not change anything in your database in case of any error or issue. Take backup & restore require more time.

Once you change the backup type, your backup chain is no longer valid. You need to take FULL backup once again and do the differential/transaction log backup as per your backup strategy.

Thanks

Jeff Moden SSC Guru Points: 1004284 More actions · Answer 8

Hardy21 (12/7/2010)
Sachin Nandanwar (12/7/2010)
Jeff Moden (12/4/2010)
Sachin Nandanwar (12/3/2010)
Also it is a better idea to put the recover model to SIMPLE or else your log file will bloat.But make sure you take a Full and a T log backup before changing the recovery model to SIMPLE.
Agh... be careful now. Changing to SIMPLE breaks the backup chain. This is normally one of the worst things you can do.
But that's the reason I suggested to take a FULL backup before changing it to SIMPLE.Then delete the records.Change it to FULL.Again take a Full Backup.
Correct me if I am missing something.
Execute the query inside the Begin transaction - End Transaction, If you satisfy with the result then commit it otherwise rollback it. So, it will not change anything in your database in case of any error or issue. Take backup & restore require more time.
Once you change the backup type, your backup chain is no longer valid. You need to take FULL backup once again and do the differential/transaction log backup as per your backup strategy.

The whole point is to try to avoid sending the LOG file through the roof with a bazillion deletes and to try to accelerate the rate of the deletions. Simply adding an explicit transaction will have neither of those effects.

--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.

Change is inevitable... Change for the better is not.

Helpful Links:
How to post code problems
How to Post Performance Problems
Create a Tally Function (fnTally)