Question

I have the following code in a cursor in SQL Server 2008 r2.

begin 
    if exists (select * from GestionIOF with (updlock,serializable) where Fecha 
    =@fechagestion and ID_Desc2 = @id_tipoalarmas and ID_Modulo = @id_modulo)
    begin 
        update GestionIOF set Total = (Select Total from #Paso p where p.Fecha =   
        @fechagestion and p.Desc2 = @id_tipoalarmas and p.Modulo = @id_modulo), 
        Frecuencia = (Select Frecuencia from #Paso p where p.Fecha 
        = @fechagestion and p.Desc2 = @id_tipoalarmas 
        and p.Modulo = @id_modulo), Gestionado = 'False'
        where GestionIOF.Fecha = @fechagestion and GestionIOF.ID_Desc2 = 
        @id_tipoalarmas and GestionIOF.ID_Modulo = @id_modulo
    end
    else
    begin
        INSERT INTO GestionIOF(ID_Area,ID_Controlador ,ID_Modulo     
        ,ID_Desc2,ID_TipoEvento,Total,Frecuencia ,Fecha,Gestionado )
        SELECT * FROM #Paso 
    end

I check if the primary key exists and then if it doesn’t I insert it. But I am getting the following error:

Mens 2627, Nivel 14, Estado 1, Procedimiento InsertarGestionEC_IOF, Línea 67
Infracción de la restricción PRIMARY KEY ‘PK_GestionIOF’. No se puede insertar una clave duplicada en el objeto ‘dbo.GestionIOF’.

Maybe it is a bug. Any ideas?

Answer 1

The exists clause is checking based on some input parameters @id_tipoalarmas and @id_modulo – but the insert clause in the else part of the statement is using #paso

I would double check what is in #paso, since that could have multiple rows, but your input parameters is just 1 set of IDS. #paso could contain other ID’s you are not checking.