Werid Behaviour of TSQL

  • I have complex view and try to fetch data from it.

    In following code query 2 takes 1 sec to execute and query 1 keeps on running for hours.

    I tried to reduce the time of query 1 to same as query 2 but no luck.I tried force order,hash join,creating clustered on temp table,type conversion etc.

    This is one of the most complicated behavior I have seen in SQL server.any idea why it is happening so

    SET statistics io ON

    SET TRANSACTION isolation level READ uncommitted

    DROP TABLE tempdb..#ps_temp

    CREATE TABLE #ps_temp

    (

    a numeric(13,0),

    b numeric(3,0)

    )

    DECLARE @a numeric(13,0),

    @b-2 numeric(3,0)

    ---fetches only 1 record

    INSERT INTO #ps_temp

    (a,

    b)

    SELECT a,

    b

    FROM tbl_x

    WHERE d = 'L140502021514'

    SELECT @a = a,

    @b-2 = b

    FROM #ps_temp

    -----------------query 1---------------------------

    SELECT t.a,

    t.b

    FROM #ps_temp d

    inner JOIN dbo.vw_complex AS t -- (hash)

    ON t.a = d.a

    AND t.b = d.b

    -- option (force order)

    -----------------query 2---------------------------

    SELECT t.a,

    t.b

    FROM #ps_temp d

    inner JOIN dbo.vw_complex AS t -- (hash)

    ON t.a = @a

    AND t.b = d.b

    Pramod
    SQL Server DBA | MCSE SQL Server 2012/2014

    in.linkedin.com/in/pramodsingla/
    http://pramodsingla.wordpress.com/

  • First a quick question, why the dirty read (READ UNCOMMITTED)?

    Without more information such as execution plans, view code, schema structure etc., it is hard to tell what the problem is.

    ๐Ÿ˜Ž

  • psingla (5/6/2014)


    I have complex view and try to fetch data from it.

    In following code query 2 takes 1 sec to execute and query 1 keeps on running for hours.

    I tried to reduce the time of query 1 to same as query 2 but no luck.I tried force order,hash join,creating clustered on temp table,type conversion etc....

    Check the execution plan of query1 for timeout. You could post the actual execution plans for both queries if you are interested in more than guesses.

    โ€œWrite the query the simplest way. If through testing it becomes clear that the performance is inadequate, consider alternative query forms.โ€ - 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

  • psingla (5/6/2014)


    I have complex view and try to fetch data from it.

    In following code query 2 takes 1 sec to execute and query 1 keeps on running for hours.

    I tried to reduce the time of query 1 to same as query 2 but no luck.I tried force order,hash join,creating clustered on temp table,type conversion etc.

    This is one of the most complicated behavior I have seen in SQL server.any idea why it is happening so

    SET statistics io ON

    SET TRANSACTION isolation level READ uncommitted

    DROP TABLE tempdb..#ps_temp

    CREATE TABLE #ps_temp

    (

    a numeric(13,0),

    b numeric(3,0)

    )

    DECLARE @a numeric(13,0),

    @b-2 numeric(3,0)

    ---fetches only 1 record

    INSERT INTO #ps_temp

    (a,

    b)

    SELECT a,

    b

    FROM tbl_x

    WHERE d = 'L140502021514'

    SELECT @a = a,

    @b-2 = b

    FROM #ps_temp

    -----------------query 1---------------------------

    SELECT t.a,

    t.b

    FROM #ps_temp d

    inner JOIN dbo.vw_complex AS t -- (hash)

    ON t.a = d.a

    AND t.b = d.b

    -- option (force order)

    -----------------query 2---------------------------

    SELECT t.a,

    t.b

    FROM #ps_temp d

    inner JOIN dbo.vw_complex AS t -- (hash)

    ON t.a = @a

    AND t.b = d.b

    Without seeing the tables or the code for the view and the fact that the setup code refers to a "tbl_x" table that doesn't seem to be available, I'd have to say that the relationship of t.a = d.a produces a many-to-many join rivaliing a Cartesian Product. Look at your execution plan for counts on arrows that are much bigger than the number of rows expected from each table.

    --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)

Viewing 4 posts - 1 through 3 (of 3 total)

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