Break down a delete statement

  • We've been using the statement below to delete rows in a certain table. The row count keeps getting larger and larger and I've been asked to break down the deletes. I need to do no more than 100,000 at a time and I need to commit every 1000. I wasn't sure if I should use a cursor or IF, Else statement.

    Here is the current statement.

     

    declare @deleteconsumerID TABLE (

    ConsumerID int NOT NULL )

     

    insert into @deleteconsumerID

    select a.consumer_id from consumer.dbo.acct a(nolock)

    left join con.dbo.con c(nolock) on a.consumer_id = c.consumer_id

    left join age.dbo.lm l(nolock) on a.consumer_id = l.consumer_id

    left join con.dbo.part_data p(nolock) on a.consumer_id = p.consumer_id

    where committedyn = '?'  and c.consumer_id is null and l.consumer_id is null

     and p.consumer_id is null and datediff(day, a.created_on, getdate()) > 365

    delete from con.dbo.acct where consumer_id in (select * from @deleteconsumerid)

  • not sure if I understand the issue correctly but would using SET ROWCOUNT be suffice in this situation.  what is the reason behind the need to break down the delete statement by 100,000 and then by 1,000 records?

  • We don't want to process more than 100,000 rows at a time because the job runs long. We need a commit every 1000 so the table is released between deletes.

    I did have the setrow count 100000 but couldn't pull the table variable into a loop to process 1000 commits.

     

  • How about:

    Declare @deleteconsumerid table (pk int identity(1,1), cons_id Int), @rows Int

    Set @rows = 1000

    While (Select count(*) From @deleteconsumerid) > 0

    Begin

    delete A from con.dbo.acct A Join @deletesconsumerid D On A.consumer_id = D.cons_id where D.pk <=@rows

    Delete from @deleteconsumerid where pk <= @rows

    Set @rows = @rows + 1000

    End

  • Thanks for your help, I tried the code above added with my insert statement, but not sure I'm declaring my table variable in the right spot. Here is the error message I received and the script I ran. Any ideas?

    Server: Msg 137, Level 15, State 2, Line 19

    Must declare the variable '@deletesconsumerid'.

     

    Declare @deleteconsumerid table (pk int identity(1,1), cons_id Int)

    declare @rows Int

    insert into @deleteconsumerID

    select a.consumer_id from con.dbo.acct a(nolock)

    left join con.dbo.con c(nolock) on a.consumer_id = c.consumer_id

    left join age.dbo.lm l(nolock) on a.consumer_id = l.consumer_id

    left join con.dbo.part_data p(nolock) on a.consumer_id = p.consumer_id

    where committedyn = '?'  and c.consumer_id is null and l.consumer_id is null

     and p.consumer_id is null and datediff(day, a.created_on, getdate()) > 365

    Set @rows = 1000

    While (Select count(*) From @deleteconsumerid) > 0

    Begin

    delete A from con.dbo.acct A Join @deletesconsumerid D On A.consumer_id = D.cons_id where D.pk <=@rows

    Delete from @deleteconsumerid where pk <= @rows

    Set @rows = @rows + 1000

    End

     

     

     

     

     

  • Typo on my part down in the delete satement.  It says @deletesconsumerid instead of @deleteconsumerid.

     

  • Thanks for all your help.

    Susan 

  • Why bother with the TABLE variable?

    declare @rows int

    set @rows = 0

    while (@rows < 100000)

    begin

       delete con.dbo.acct

       from (

          select TOP 1000 a.consumer_id  as [ID]

          from

          consumer.dbo.acct a(nolock)

             left join con.dbo.con c(nolock)

             on a.consumer_id = c.consumer_id

                left join age.dbo.lm l(nolock)

                on a.consumer_id = l.consumer_id

                   left join con.dbo.part_data p(nolock)

                   on a.consumer_id = p.consumer_id

          where

              committedyn = '?' 

          and c.consumer_id is null

          and l.consumer_id is null

          and p.consumer_id is null

          and datediff(day, a.created_on, getdate()) > 365

          ) as [RANGE]

       where

          con.dbo.acct.consumer_id = [RANGE].[ID]

       set @rows = @rows + 1000

    end

    Peter Evans (__PETER Peter_)

  • how do I apply this same solution to commit every 1000 inserts??

  • The example is commiting every 1000 records.

    My requirments were bring in no more than 100,000 rows then execute the inserts 1000 at a time.

    Hopefully that's what you were referring to.

    Thanks

    Susan

     

Viewing 10 posts - 1 through 9 (of 9 total)

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