I have the following query returning results in opposite order between
versions 2000 and 2005:-
SELECT * FROM tblname where form= 'L'
order by acolumn desc
The database has been migrated from 2000 to 2005, so the structure and data
should be identical. Here's the DDL for the table in SQL2000:-
USE [MSD_Contracts]
GO
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
SET ANSI_PADDING ON
GO
CREATE TABLE [dbo].[tblname](
[Form] [varchar](1) NOT NULL,
[acolumn] [varchar](1) NOT NULL,
[FieldNo] [int] NOT NULL,
[dbField] [varchar](30) NOT NULL,
[Validation] [varchar](20) NOT NULL,
[AllowNulls] [varchar](1) NOT NULL,
CONSTRAINT [PK_tblname_1_12] PRIMARY KEY CLUSTERED
(
[FieldNo] ASC,
[Form] ASC,
[acolumn] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON
[PRIMARY]
) ON [PRIMARY]
GO
SET ANSI_PADDING OFF
Any ideas please?Hi Adrian
The order should be the case assuming that the collations are the same. Do
you have some sample data? http://www.aspfaq.com/etiquette.asp?id=5006
You should change the varchar(1) columns.
John
"Adrian" wrote:
> I have the following query returning results in opposite order between
> versions 2000 and 2005:-
> SELECT * FROM tblname where form= 'L'
> order by acolumn desc
> The database has been migrated from 2000 to 2005, so the structure and data
> should be identical. Here's the DDL for the table in SQL2000:-
> USE [MSD_Contracts]
> GO
> SET ANSI_NULLS ON
> GO
> SET QUOTED_IDENTIFIER ON
> GO
> SET ANSI_PADDING ON
> GO
> CREATE TABLE [dbo].[tblname](
> [Form] [varchar](1) NOT NULL,
> [acolumn] [varchar](1) NOT NULL,
> [FieldNo] [int] NOT NULL,
> [dbField] [varchar](30) NOT NULL,
> [Validation] [varchar](20) NOT NULL,
> [AllowNulls] [varchar](1) NOT NULL,
> CONSTRAINT [PK_tblname_1_12] PRIMARY KEY CLUSTERED
> (
> [FieldNo] ASC,
> [Form] ASC,
> [acolumn] ASC
> )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY => OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON
> [PRIMARY]
> ) ON [PRIMARY]
> GO
> SET ANSI_PADDING OFF
> Any ideas please?|||Collation sequence is the same in both versions -
SQL_Latin1_General_CP1_CI_AS. Here is some sample data:_
Form FrontBack FieldNo dbField Validation
AllowNulls
-- -- -- --
-- --
L F 470 SignaturePresent AlphaNum
N
L F 210 DependentQ Bit
N
L F 200 DependentN Bit
N
L F 190 DependentY Bit
N
L F 180 MarriageCertQ Bit
N
L F 170 MarriageCertN Bit
N
L F 160 MarriageCertY Bit
N
L F 150 ConsentQ Bit
N
L F 140 ConsentN Bit
N
L F 130 ConsentY Bit
N
L F 120 Under18Years Bit
N
L F 110 CAltered Bit
N
L F 100 SignedAndDatedQ Bit
N
L F 90 SignedAndDatedN Bit
N
L F 80 SignedAndDatedY Bit
N
L F 70 NameAndDOBQ Bit
N
L F 60 NameAndDOBN Bit
N
L F 50 NameAndDOBY Bit
N
L F 40 Barcode Num
N
L F 30 FormType1 AlphaNum
N
L F 20 ImageRef1 AlphaNum
N
L F 10 BatchNo Num
N
L B 420 EvidenceSent Bit
N
L B 410 EvidenceToWinz Bit
N
L B 400 ExtraEvidence Bit
N
L B 390 Oaltered Bit
N
L B 380 BankQ Bit
N
L B 370 BankN Bit
N
L B 360 BankY Bit
N
L B 340 IRDQ Bit
N
L B 330 IRDN Bit
N
L B 320 IRDY Bit
N
L B 310 ResidentQ Bit
N
L B 300 ResidentN Bit
N
L B 290 ResidentY Bit
N
L B 280 CitizenQ Bit
N
L B 270 CitizenN Bit
N
L B 260 CitizenY Bit
N
L B 250 Barcode Num
N
L B 240 FormType2 AlphaNum
N
L B 230 ImageRef2 AlphaNum
N
L B 220 BatchNo Num
N
(42 row(s) affected)
"John Bell" wrote:
> Hi Adrian
> The order should be the case assuming that the collations are the same. Do
> you have some sample data? http://www.aspfaq.com/etiquette.asp?id=5006
> You should change the varchar(1) columns.
> John
> "Adrian" wrote:
> > I have the following query returning results in opposite order between
> > versions 2000 and 2005:-
> > SELECT * FROM tblname where form= 'L'
> > order by acolumn desc
> >
> > The database has been migrated from 2000 to 2005, so the structure and data
> > should be identical. Here's the DDL for the table in SQL2000:-
> > USE [MSD_Contracts]
> > GO
> > SET ANSI_NULLS ON
> > GO
> > SET QUOTED_IDENTIFIER ON
> > GO
> > SET ANSI_PADDING ON
> > GO
> > CREATE TABLE [dbo].[tblname](
> > [Form] [varchar](1) NOT NULL,
> > [acolumn] [varchar](1) NOT NULL,
> > [FieldNo] [int] NOT NULL,
> > [dbField] [varchar](30) NOT NULL,
> > [Validation] [varchar](20) NOT NULL,
> > [AllowNulls] [varchar](1) NOT NULL,
> > CONSTRAINT [PK_tblname_1_12] PRIMARY KEY CLUSTERED
> > (
> > [FieldNo] ASC,
> > [Form] ASC,
> > [acolumn] ASC
> > )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY => > OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON
> > [PRIMARY]
> > ) ON [PRIMARY]
> >
> > GO
> > SET ANSI_PADDING OFF
> >
> > Any ideas please?|||Hi Arian
You did not read the link I posted about sample data.
Here is some code that would have been useful:
USE TEMPDB
GO
CREATE TABLE [dbo].[tblname](
[Form] [varchar](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[FrontBack] [varchar](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[FieldNo] [int] NOT NULL ,
[dbField] [varchar](30) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[Validation] [varchar](20) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[AllowNulls] [varchar](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
CONSTRAINT [PK_tblname_1_12] PRIMARY KEY CLUSTERED
(
[FieldNo] ASC,
[Form] ASC,
[FrontBack] ASC
)
)
INSERT INTO tblname (Form, FrontBack, FieldNo, dbField, Validation,
AllowNulls)
SELECT 'L','F',470,'SignaturePresent','AlphaNum','N'
UNION ALL SELECT 'L','F',210,'DependentQ','Bit','N'
UNION ALL SELECT 'L','F',200,'DependentN','Bit','N'
UNION ALL SELECT 'L','F',190,'DependentY','Bit','N'
UNION ALL SELECT 'L','F',180,'MarriageCertQ','Bit','N'
UNION ALL SELECT 'L','F',170,'MarriageCertN','Bit','N'
UNION ALL SELECT 'L','F',160,'MarriageCertY','Bit','N'
UNION ALL SELECT 'L','F',150,'ConsentQ','Bit','N'
UNION ALL SELECT 'L','F',140,'ConsentN','Bit','N'
UNION ALL SELECT 'L','F',130,'ConsentY','Bit','N'
UNION ALL SELECT 'L','F',120,'Under18Years','Bit','N'
UNION ALL SELECT 'L','F',110,'CAltered','Bit','N'
UNION ALL SELECT 'L','F',100,'SignedAndDatedQ','Bit','N'
UNION ALL SELECT 'L','F',90,'SignedAndDatedN','Bit','N'
UNION ALL SELECT 'L','F',80,'SignedAndDatedY','Bit','N'
UNION ALL SELECT 'L','F',70,'NameAndDOBQ','Bit','N'
UNION ALL SELECT 'L','F',60,'NameAndDOBN','Bit','N'
UNION ALL SELECT 'L','F',50,'NameAndDOBY','Bit','N'
UNION ALL SELECT 'L','F',40,'Barcode','Num','N'
UNION ALL SELECT 'L','F',30,'FormType1','AlphaNum','N'
UNION ALL SELECT 'L','F',20,'ImageRef1','AlphaNum','N'
UNION ALL SELECT 'L','F',10,'BatchNo','Num','N'
UNION ALL SELECT 'L','B',420,'EvidenceSent','Bit','N'
UNION ALL SELECT 'L','B',410,'EvidenceToWinz','Bit','N'
UNION ALL SELECT 'L','B',400,'ExtraEvidence','Bit','N'
UNION ALL SELECT 'L','B',390,'Oaltered','Bit','N'
UNION ALL SELECT 'L','B',380,'BankQ','Bit','N'
UNION ALL SELECT 'L','B',370,'BankN','Bit','N'
UNION ALL SELECT 'L','B',360,'BankY','Bit','N'
UNION ALL SELECT 'L','B',340,'IRDQ','Bit','N'
UNION ALL SELECT 'L','B',330,'IRDN','Bit','N'
UNION ALL SELECT 'L','B',320,'IRDY','Bit','N'
UNION ALL SELECT 'L','B',310,'ResidentQ','Bit','N'
UNION ALL SELECT 'L','B',300,'ResidentN','Bit','N'
UNION ALL SELECT 'L','B',290,'ResidentY','Bit','N'
UNION ALL SELECT 'L','B',280,'CitizenQ','Bit','N'
UNION ALL SELECT 'L','B',270,'CitizenN','Bit','N'
UNION ALL SELECT 'L','B',260,'CitizenY','Bit','N'
UNION ALL SELECT 'L','B',250,'Barcode','Num','N'
UNION ALL SELECT 'L','B',240,'FormType2','AlphaNum','N'
UNION ALL SELECT 'L','B',230,'ImageRef2','AlphaNum','N'
UNION ALL SELECT 'L','B',220,'BatchNo','Num','N'
SELECT *
FROM tblname
where form= 'L'
order by FrontBack desc
The output I get from SQL 2000 is:
Form FrontBack FieldNo dbField Validation
AllowNulls
-- -- -- --
-- --
L F 10 BatchNo Num
N
L F 20 ImageRef1 AlphaNum
N
L F 30 FormType1 AlphaNum
N
L F 40 Barcode Num
N
L F 50 NameAndDOBY Bit
N
L F 60 NameAndDOBN Bit
N
L F 70 NameAndDOBQ Bit
N
L F 80 SignedAndDatedY Bit
N
L F 90 SignedAndDatedN Bit
N
L F 100 SignedAndDatedQ Bit
N
L F 110 CAltered Bit
N
L F 120 Under18Years Bit
N
L F 130 ConsentY Bit
N
L F 140 ConsentN Bit
N
L F 150 ConsentQ Bit
N
L F 160 MarriageCertY Bit
N
L F 170 MarriageCertN Bit
N
L F 180 MarriageCertQ Bit
N
L F 190 DependentY Bit
N
L F 200 DependentN Bit
N
L F 210 DependentQ Bit
N
L F 470 SignaturePresent AlphaNum
N
L B 220 BatchNo Num
N
L B 230 ImageRef2 AlphaNum
N
L B 240 FormType2 AlphaNum
N
L B 250 Barcode Num
N
L B 260 CitizenY Bit
N
L B 270 CitizenN Bit
N
L B 280 CitizenQ Bit
N
L B 290 ResidentY Bit
N
L B 300 ResidentN Bit
N
L B 310 ResidentQ Bit
N
L B 320 IRDY Bit
N
L B 330 IRDN Bit
N
L B 340 IRDQ Bit
N
L B 360 BankY Bit
N
L B 370 BankN Bit
N
L B 380 BankQ Bit
N
L B 390 Oaltered Bit
N
L B 400 ExtraEvidence Bit
N
L B 410 EvidenceToWinz Bit
N
L B 420 EvidenceSent Bit
N
(42 row(s) affected)
The output I get from SQL 2005 is:
Form FrontBack FieldNo dbField Validation
AllowNulls
-- -- -- --
-- --
L F 10 BatchNo Num
N
L F 20 ImageRef1 AlphaNum
N
L F 30 FormType1 AlphaNum
N
L F 40 Barcode Num
N
L F 50 NameAndDOBY Bit
N
L F 60 NameAndDOBN Bit
N
L F 70 NameAndDOBQ Bit
N
L F 80 SignedAndDatedY Bit
N
L F 90 SignedAndDatedN Bit
N
L F 100 SignedAndDatedQ Bit
N
L F 110 CAltered Bit
N
L F 120 Under18Years Bit
N
L F 130 ConsentY Bit
N
L F 140 ConsentN Bit
N
L F 150 ConsentQ Bit
N
L F 160 MarriageCertY Bit
N
L F 170 MarriageCertN Bit
N
L F 180 MarriageCertQ Bit
N
L F 190 DependentY Bit
N
L F 200 DependentN Bit
N
L F 210 DependentQ Bit
N
L F 470 SignaturePresent AlphaNum
N
L B 220 BatchNo Num
N
L B 230 ImageRef2 AlphaNum
N
L B 240 FormType2 AlphaNum
N
L B 250 Barcode Num
N
L B 260 CitizenY Bit
N
L B 270 CitizenN Bit
N
L B 280 CitizenQ Bit
N
L B 290 ResidentY Bit
N
L B 300 ResidentN Bit
N
L B 310 ResidentQ Bit
N
L B 320 IRDY Bit
N
L B 330 IRDN Bit
N
L B 340 IRDQ Bit
N
L B 360 BankY Bit
N
L B 370 BankN Bit
N
L B 380 BankQ Bit
N
L B 390 Oaltered Bit
N
L B 400 ExtraEvidence Bit
N
L B 410 EvidenceToWinz Bit
N
L B 420 EvidenceSent Bit
N
(42 row(s) affected)
The order by only guarantees that FrontBack will be ordered 'F' then 'B' and
this is the case on both instances. As FieldNo and Form are not in the order
by their order is not guaranteed.
John
"Adrian" wrote:
> Collation sequence is the same in both versions -
> SQL_Latin1_General_CP1_CI_AS. Here is some sample data:_
> Form FrontBack FieldNo dbField Validation
> AllowNulls
> -- -- -- --
> -- --
> L F 470 SignaturePresent AlphaNum
> N
> L F 210 DependentQ Bit
> N
> L F 200 DependentN Bit
> N
> L F 190 DependentY Bit
> N
> L F 180 MarriageCertQ Bit
> N
> L F 170 MarriageCertN Bit
> N
> L F 160 MarriageCertY Bit
> N
> L F 150 ConsentQ Bit
> N
> L F 140 ConsentN Bit
> N
> L F 130 ConsentY Bit
> N
> L F 120 Under18Years Bit
> N
> L F 110 CAltered Bit
> N
> L F 100 SignedAndDatedQ Bit
> N
> L F 90 SignedAndDatedN Bit
> N
> L F 80 SignedAndDatedY Bit
> N
> L F 70 NameAndDOBQ Bit
> N
> L F 60 NameAndDOBN Bit
> N
> L F 50 NameAndDOBY Bit
> N
> L F 40 Barcode Num
> N
> L F 30 FormType1 AlphaNum
> N
> L F 20 ImageRef1 AlphaNum
> N
> L F 10 BatchNo Num
> N
> L B 420 EvidenceSent Bit
> N
> L B 410 EvidenceToWinz Bit
> N
> L B 400 ExtraEvidence Bit
> N
> L B 390 Oaltered Bit
> N
> L B 380 BankQ Bit
> N
> L B 370 BankN Bit
> N
> L B 360 BankY Bit
> N
> L B 340 IRDQ Bit
> N
> L B 330 IRDN Bit
> N
> L B 320 IRDY Bit
> N
> L B 310 ResidentQ Bit
> N
> L B 300 ResidentN Bit
> N
> L B 290 ResidentY Bit
> N
> L B 280 CitizenQ Bit
> N
> L B 270 CitizenN Bit
> N
> L B 260 CitizenY Bit
> N
> L B 250 Barcode Num
> N
> L B 240 FormType2 AlphaNum
> N
> L B 230 ImageRef2 AlphaNum
> N
> L B 220 BatchNo Num
> N
> (42 row(s) affected)
>
> "John Bell" wrote:
> > Hi Adrian
> >
> > The order should be the case assuming that the collations are the same. Do
> > you have some sample data? http://www.aspfaq.com/etiquette.asp?id=5006
> >
> > You should change the varchar(1) columns.
> >
> > John
> >
> > "Adrian" wrote:
> >
> > > I have the following query returning results in opposite order between
> > > versions 2000 and 2005:-
> > > SELECT * FROM tblname where form= 'L'
> > > order by acolumn desc
> > >
> > > The database has been migrated from 2000 to 2005, so the structure and data
> > > should be identical. Here's the DDL for the table in SQL2000:-
> > > USE [MSD_Contracts]
> > > GO
> > > SET ANSI_NULLS ON
> > > GO
> > > SET QUOTED_IDENTIFIER ON
> > > GO
> > > SET ANSI_PADDING ON
> > > GO
> > > CREATE TABLE [dbo].[tblname](
> > > [Form] [varchar](1) NOT NULL,
> > > [acolumn] [varchar](1) NOT NULL,
> > > [FieldNo] [int] NOT NULL,
> > > [dbField] [varchar](30) NOT NULL,
> > > [Validation] [varchar](20) NOT NULL,
> > > [AllowNulls] [varchar](1) NOT NULL,
> > > CONSTRAINT [PK_tblname_1_12] PRIMARY KEY CLUSTERED
> > > (
> > > [FieldNo] ASC,
> > > [Form] ASC,
> > > [acolumn] ASC
> > > )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY => > > OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON, FILLFACTOR = 90) ON
> > > [PRIMARY]
> > > ) ON [PRIMARY]
> > >
> > > GO
> > > SET ANSI_PADDING OFF
> > >
> > > Any ideas please?
Showing posts with label returning. Show all posts
Showing posts with label returning. Show all posts
Wednesday, March 28, 2012
ORDER BY NEWID() returning duplicates records
This returns 2 random records for me. However, on occasion, the 2 records
returned are identical. Does anybody have any thought for preventing this
from happening?
SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
C WHERE P.Image1<>'' AND C.ProductID = P.ProductID ORDER BY NEWID()
TIA,
TonyGThat's because SQL evaluates NewId() only once in your query. If you are try
ing
to randomly return two records try this:
Select TOP2 NewId(), P.*, C.CatId
From JEP_tblProducts As P
, JEP_LtblProductCategories As C
Where P.Image1 <> ''
And C.ProductId = P.ProductId
Order By 1
Thomas|||Hi
It could be because the join results in 2 identical records in some instance
s.
Run the query with no TOP clause and see how that compares to the number of
records in the tables.
You may want to update your query to the ANSI style join:
SELECT
TOP 2
P.*,
C.CatID
FROM
JEP_tblProducts AS P
INNER JOJN JEP_LtblProductCategories AS C
ON C.ProductID = P.ProductID
WHERE
P.Image1<>''
ORDER BY NEWID()
Regards
Mike
"TonyG" wrote:
> This returns 2 random records for me. However, on occasion, the 2 records
> returned are identical. Does anybody have any thought for preventing this
> from happening?
> SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategorie
s
> C WHERE P.Image1<>'' AND C.ProductID = P.ProductID ORDER BY NEWID()
> TIA,
> TonyG
>
>|||>> That's because SQL evaluates NewId() only once in your query.
Generally, any non-numeric expression used in the ORDER BY clause is
evaluated in the same way the expression is evaluated in the SELECT clause.
And it is applicable for NEWID() as well.
The duplication OP has is perhaps due to lack of keys/constraints in the
tables and/or the result of using non-unique columns in the JOIN clause. But
then without seeing any useful DDLs, it would be anyone's guess.
Anith|||Thats not true, using the newid () funtion in the order generates for every
row in the query a seperate GUID , every time you execute is. There is a
problem with your query it returns more than one row becasue your table in
not only related to the second table via one column.
Try to select this and youll see that there will be more than one row
returned:
Select count(*) from
> From JEP_tblProducts As P
> , JEP_LtblProductCategories As C
> Where P.Image1 <> ''
> And C.ProductId = P.ProductId
Perhaps you can make a join with the appropiate key in it.
HTH, Jens Suessmeyer.
http://www.sqlserver2005.de
--
"Thomas" <thomas@.newsgroup.nospam> schrieb im Newsbeitrag
news:Onzdy%23nSFHA.248@.TK2MSFTNGP15.phx.gbl...
> That's because SQL evaluates NewId() only once in your query. If you are
> trying to randomly return two records try this:
> Select TOP2 NewId(), P.*, C.CatId
> From JEP_tblProducts As P
> , JEP_LtblProductCategories As C
> Where P.Image1 <> ''
> And C.ProductId = P.ProductId
> Order By 1
>
> Thomas
>|||Well, you have a 1-M relationship. Thus, it's possible to return 2 that are
identical. The newid() only randomizes the resultset and does not remove the
dupes.
Here is a trick to get the uniqueness.
SELECT TOP 2 *
from (select distinct
P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
C WHERE P.Image1<>'' AND C.ProductID = P.ProductID
) derived
ORDER BY NEWID()
-oj
"TonyG" <groups@.RE-MO-VE-globalmagic.com> wrote in message
news:OUBU05nSFHA.3088@.TK2MSFTNGP15.phx.gbl...
> This returns 2 random records for me. However, on occasion, the 2 records
> returned are identical. Does anybody have any thought for preventing this
> from happening?
> SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P,
> JEP_LtblProductCategories C WHERE P.Image1<>'' AND C.ProductID =
> P.ProductID ORDER BY NEWID()
> TIA,
> TonyG
>|||I stand corrected. Is there is any way to know ths for sure (documentation,
proof of concept etc)?
Thomas
"Anith Sen" <anith@.bizdatasolutions.com> wrote in message
news:eGCxkEoSFHA.1396@.TK2MSFTNGP10.phx.gbl...
> Generally, any non-numeric expression used in the ORDER BY clause is evalu
ated
> in the same way the expression is evaluated in the SELECT clause. And it i
s
> applicable for NEWID() as well.
> The duplication OP has is perhaps due to lack of keys/constraints in the
> tables and/or the result of using non-unique columns in the JOIN clause. B
ut
> then without seeing any useful DDLs, it would be anyone's guess.
> --
> Anith
>|||Build an example myself. Thanks for the correction.
Thomas
"Anith Sen" <anith@.bizdatasolutions.com> wrote in message
news:eGCxkEoSFHA.1396@.TK2MSFTNGP10.phx.gbl...
> Generally, any non-numeric expression used in the ORDER BY clause is evalu
ated
> in the same way the expression is evaluated in the SELECT clause. And it i
s
> applicable for NEWID() as well.
> The duplication OP has is perhaps due to lack of keys/constraints in the
> tables and/or the result of using non-unique columns in the JOIN clause. B
ut
> then without seeing any useful DDLs, it would be anyone's guess.
> --
> Anith
>|||Thanks, that did the trick.
I knew I was missing something.
TonyG
"oj" <nospam_ojngo@.home.com> wrote in message
news:OJSm6JoSFHA.2000@.TK2MSFTNGP10.phx.gbl...
> Well, you have a 1-M relationship. Thus, it's possible to return 2 that
> are identical. The newid() only randomizes the resultset and does not
> remove the dupes.
> Here is a trick to get the uniqueness.
> SELECT TOP 2 *
> from (select distinct
> P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
> C WHERE P.Image1<>'' AND C.ProductID = P.ProductID
> ) derived
> ORDER BY NEWID()
> --
> -oj
>
> "TonyG" <groups@.RE-MO-VE-globalmagic.com> wrote in message
> news:OUBU05nSFHA.3088@.TK2MSFTNGP15.phx.gbl...
>
returned are identical. Does anybody have any thought for preventing this
from happening?
SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
C WHERE P.Image1<>'' AND C.ProductID = P.ProductID ORDER BY NEWID()
TIA,
TonyGThat's because SQL evaluates NewId() only once in your query. If you are try
ing
to randomly return two records try this:
Select TOP2 NewId(), P.*, C.CatId
From JEP_tblProducts As P
, JEP_LtblProductCategories As C
Where P.Image1 <> ''
And C.ProductId = P.ProductId
Order By 1
Thomas|||Hi
It could be because the join results in 2 identical records in some instance
s.
Run the query with no TOP clause and see how that compares to the number of
records in the tables.
You may want to update your query to the ANSI style join:
SELECT
TOP 2
P.*,
C.CatID
FROM
JEP_tblProducts AS P
INNER JOJN JEP_LtblProductCategories AS C
ON C.ProductID = P.ProductID
WHERE
P.Image1<>''
ORDER BY NEWID()
Regards
Mike
"TonyG" wrote:
> This returns 2 random records for me. However, on occasion, the 2 records
> returned are identical. Does anybody have any thought for preventing this
> from happening?
> SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategorie
s
> C WHERE P.Image1<>'' AND C.ProductID = P.ProductID ORDER BY NEWID()
> TIA,
> TonyG
>
>|||>> That's because SQL evaluates NewId() only once in your query.
Generally, any non-numeric expression used in the ORDER BY clause is
evaluated in the same way the expression is evaluated in the SELECT clause.
And it is applicable for NEWID() as well.
The duplication OP has is perhaps due to lack of keys/constraints in the
tables and/or the result of using non-unique columns in the JOIN clause. But
then without seeing any useful DDLs, it would be anyone's guess.
Anith|||Thats not true, using the newid () funtion in the order generates for every
row in the query a seperate GUID , every time you execute is. There is a
problem with your query it returns more than one row becasue your table in
not only related to the second table via one column.
Try to select this and youll see that there will be more than one row
returned:
Select count(*) from
> From JEP_tblProducts As P
> , JEP_LtblProductCategories As C
> Where P.Image1 <> ''
> And C.ProductId = P.ProductId
Perhaps you can make a join with the appropiate key in it.
HTH, Jens Suessmeyer.
http://www.sqlserver2005.de
--
"Thomas" <thomas@.newsgroup.nospam> schrieb im Newsbeitrag
news:Onzdy%23nSFHA.248@.TK2MSFTNGP15.phx.gbl...
> That's because SQL evaluates NewId() only once in your query. If you are
> trying to randomly return two records try this:
> Select TOP2 NewId(), P.*, C.CatId
> From JEP_tblProducts As P
> , JEP_LtblProductCategories As C
> Where P.Image1 <> ''
> And C.ProductId = P.ProductId
> Order By 1
>
> Thomas
>|||Well, you have a 1-M relationship. Thus, it's possible to return 2 that are
identical. The newid() only randomizes the resultset and does not remove the
dupes.
Here is a trick to get the uniqueness.
SELECT TOP 2 *
from (select distinct
P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
C WHERE P.Image1<>'' AND C.ProductID = P.ProductID
) derived
ORDER BY NEWID()
-oj
"TonyG" <groups@.RE-MO-VE-globalmagic.com> wrote in message
news:OUBU05nSFHA.3088@.TK2MSFTNGP15.phx.gbl...
> This returns 2 random records for me. However, on occasion, the 2 records
> returned are identical. Does anybody have any thought for preventing this
> from happening?
> SELECT TOP 2 P.*, C.CatID FROM JEP_tblProducts P,
> JEP_LtblProductCategories C WHERE P.Image1<>'' AND C.ProductID =
> P.ProductID ORDER BY NEWID()
> TIA,
> TonyG
>|||I stand corrected. Is there is any way to know ths for sure (documentation,
proof of concept etc)?
Thomas
"Anith Sen" <anith@.bizdatasolutions.com> wrote in message
news:eGCxkEoSFHA.1396@.TK2MSFTNGP10.phx.gbl...
> Generally, any non-numeric expression used in the ORDER BY clause is evalu
ated
> in the same way the expression is evaluated in the SELECT clause. And it i
s
> applicable for NEWID() as well.
> The duplication OP has is perhaps due to lack of keys/constraints in the
> tables and/or the result of using non-unique columns in the JOIN clause. B
ut
> then without seeing any useful DDLs, it would be anyone's guess.
> --
> Anith
>|||Build an example myself. Thanks for the correction.
Thomas
"Anith Sen" <anith@.bizdatasolutions.com> wrote in message
news:eGCxkEoSFHA.1396@.TK2MSFTNGP10.phx.gbl...
> Generally, any non-numeric expression used in the ORDER BY clause is evalu
ated
> in the same way the expression is evaluated in the SELECT clause. And it i
s
> applicable for NEWID() as well.
> The duplication OP has is perhaps due to lack of keys/constraints in the
> tables and/or the result of using non-unique columns in the JOIN clause. B
ut
> then without seeing any useful DDLs, it would be anyone's guess.
> --
> Anith
>|||Thanks, that did the trick.
I knew I was missing something.
TonyG
"oj" <nospam_ojngo@.home.com> wrote in message
news:OJSm6JoSFHA.2000@.TK2MSFTNGP10.phx.gbl...
> Well, you have a 1-M relationship. Thus, it's possible to return 2 that
> are identical. The newid() only randomizes the resultset and does not
> remove the dupes.
> Here is a trick to get the uniqueness.
> SELECT TOP 2 *
> from (select distinct
> P.*, C.CatID FROM JEP_tblProducts P, JEP_LtblProductCategories
> C WHERE P.Image1<>'' AND C.ProductID = P.ProductID
> ) derived
> ORDER BY NEWID()
> --
> -oj
>
> "TonyG" <groups@.RE-MO-VE-globalmagic.com> wrote in message
> news:OUBU05nSFHA.3088@.TK2MSFTNGP15.phx.gbl...
>
Monday, March 26, 2012
ORDER BY is very slow when querying linked Server Oracle9i
I am creating 3 views from the linked server
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal '
code below
--
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEWHi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal '
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>
>
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal '
code below
--
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEWHi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal '
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>
>
ORDER BY is very slow when querying linked Server Oracle9i
I am creating 3 views from the linked server
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal ?
code below
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEW
Hi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal ?
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>
|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>
>
sql
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal ?
code below
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEW
Hi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal ?
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>
|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>
>
sql
ORDER BY is very slow when querying linked Server Oracle9i
I am creating 3 views from the linked server
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal '
code below
--
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEWHi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal '
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
> >I am creating 3 views from the linked server
> > and then I have a question at the bottom returning results.
> > If I exclude the ORDER BY from my question the query will take about 40
> > seconds,
> > with the ORDER BY it takes about an hour to finish.
> >
> > Is this normal '
> >
> > code below
> > --
> >
> > create view xxxyyy1 as
> > SELECT INSATSID, STATUS
> > FROM HACTAR..EKOP2.INSATS
> >
> > go
> >
> > create view xxxyyy2 as
> > SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> >
> > go
> >
> > create view xxxyyy3 as
> > select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> > 'INK905_STATUS'
> > from xxxyyy1 A left outer join xxxyyy2 B
> > on a.insatsid = b.insats
> > where a.status <> b.status
> > go
> >
> > SELECT
> > A.INSATSID,
> > A.STATUS_OLD,
> > A.STATUS_NEW,
> > A.ANDR_DATUM,
> > A.ANDR_TID,
> > B.INSATSID,
> > B.STATUS,
> > B.INK905_INSATS,
> > B.INK905_STATUS
> > FROM HACTAR..EKOP2.INSATSLOG A
> > right outer join xxxyyy3 B
> > on A.INSATSID = B.INSATSID
> > WHERE A.ANDR_DATUM < 20060225
> > ORDER BY
> > A.INSATSID,
> > A.ANDR_DATUM,
> > A.ANDR_TID,
> > A.STATUS_OLD,
> > A.STATUS_NEW
> >
>
>
and then I have a question at the bottom returning results.
If I exclude the ORDER BY from my question the query will take about 40
seconds,
with the ORDER BY it takes about an hour to finish.
Is this normal '
code below
--
create view xxxyyy1 as
SELECT INSATSID, STATUS
FROM HACTAR..EKOP2.INSATS
go
create view xxxyyy2 as
SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
go
create view xxxyyy3 as
select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
'INK905_STATUS'
from xxxyyy1 A left outer join xxxyyy2 B
on a.insatsid = b.insats
where a.status <> b.status
go
SELECT
A.INSATSID,
A.STATUS_OLD,
A.STATUS_NEW,
A.ANDR_DATUM,
A.ANDR_TID,
B.INSATSID,
B.STATUS,
B.INK905_INSATS,
B.INK905_STATUS
FROM HACTAR..EKOP2.INSATSLOG A
right outer join xxxyyy3 B
on A.INSATSID = B.INSATSID
WHERE A.ANDR_DATUM < 20060225
ORDER BY
A.INSATSID,
A.ANDR_DATUM,
A.ANDR_TID,
A.STATUS_OLD,
A.STATUS_NEWHi
See an execution plan. My guess in order to increase performance you may
want to add an index on the column in ORDER BY clause
"ln54" <ln54@.discussions.microsoft.com> wrote in message
news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
>I am creating 3 views from the linked server
> and then I have a question at the bottom returning results.
> If I exclude the ORDER BY from my question the query will take about 40
> seconds,
> with the ORDER BY it takes about an hour to finish.
> Is this normal '
> code below
> --
> create view xxxyyy1 as
> SELECT INSATSID, STATUS
> FROM HACTAR..EKOP2.INSATS
> go
> create view xxxyyy2 as
> SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> go
> create view xxxyyy3 as
> select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> 'INK905_STATUS'
> from xxxyyy1 A left outer join xxxyyy2 B
> on a.insatsid = b.insats
> where a.status <> b.status
> go
> SELECT
> A.INSATSID,
> A.STATUS_OLD,
> A.STATUS_NEW,
> A.ANDR_DATUM,
> A.ANDR_TID,
> B.INSATSID,
> B.STATUS,
> B.INK905_INSATS,
> B.INK905_STATUS
> FROM HACTAR..EKOP2.INSATSLOG A
> right outer join xxxyyy3 B
> on A.INSATSID = B.INSATSID
> WHERE A.ANDR_DATUM < 20060225
> ORDER BY
> A.INSATSID,
> A.ANDR_DATUM,
> A.ANDR_TID,
> A.STATUS_OLD,
> A.STATUS_NEW
>|||Thankyou, the ORDER BY statement causes it to do a lot of Nested loops
which seems to be very time consuming
"Uri Dimant" wrote:
> Hi
> See an execution plan. My guess in order to increase performance you may
> want to add an index on the column in ORDER BY clause
>
>
>
> "ln54" <ln54@.discussions.microsoft.com> wrote in message
> news:7B8F4D82-6F9B-423A-A3A4-EC957A72A876@.microsoft.com...
> >I am creating 3 views from the linked server
> > and then I have a question at the bottom returning results.
> > If I exclude the ORDER BY from my question the query will take about 40
> > seconds,
> > with the ORDER BY it takes about an hour to finish.
> >
> > Is this normal '
> >
> > code below
> > --
> >
> > create view xxxyyy1 as
> > SELECT INSATSID, STATUS
> > FROM HACTAR..EKOP2.INSATS
> >
> > go
> >
> > create view xxxyyy2 as
> > SELECT distinct insats, status FROM HACTAR..EKOP2.RVAINK001905
> >
> > go
> >
> > create view xxxyyy3 as
> > select a.insatsid, a.status, b.insats AS 'INK905_INSATS', b.status AS
> > 'INK905_STATUS'
> > from xxxyyy1 A left outer join xxxyyy2 B
> > on a.insatsid = b.insats
> > where a.status <> b.status
> > go
> >
> > SELECT
> > A.INSATSID,
> > A.STATUS_OLD,
> > A.STATUS_NEW,
> > A.ANDR_DATUM,
> > A.ANDR_TID,
> > B.INSATSID,
> > B.STATUS,
> > B.INK905_INSATS,
> > B.INK905_STATUS
> > FROM HACTAR..EKOP2.INSATSLOG A
> > right outer join xxxyyy3 B
> > on A.INSATSID = B.INSATSID
> > WHERE A.ANDR_DATUM < 20060225
> > ORDER BY
> > A.INSATSID,
> > A.ANDR_DATUM,
> > A.ANDR_TID,
> > A.STATUS_OLD,
> > A.STATUS_NEW
> >
>
>
Monday, March 12, 2012
Oracle report returning strange data
I have a RS report that is connecting to an Oracle Database (8.1.7), the
8.1.7 oracle client is installed on RS Sever, and the result set is returning
strange data. I'm using a simple view that was created in the Oracle
database, that returns a small record set about 100 records. I'm seeing ASCII
values in place of some data, in the Data tab and the Preview tab and the
Published report.
I have tried using both Oracle Drivers and the Microsoft ODBC drivers for
Oracle, and I get the same results.
Some of the data looks just fine but some of the characters get replaced by
ASCII values.
I have installed RS SP1.
Don't know what I am doing wrong...Hi Gary
We are running SQL Reporting Services (SP1) against an Oracle 9i database
without any problems. In our case 9.2.0.4 client is used to talk with
9.2.0.4 database.
I suspect the problem you are having is related to version of Oracle client
you are using. If you have access to a client 9.2.0.4 or higher you may want
to give that a try.
"garyc" wrote:
> I have a RS report that is connecting to an Oracle Database (8.1.7), the
> 8.1.7 oracle client is installed on RS Sever, and the result set is returning
> strange data. I'm using a simple view that was created in the Oracle
> database, that returns a small record set about 100 records. I'm seeing ASCII
> values in place of some data, in the Data tab and the Preview tab and the
> Published report.
> I have tried using both Oracle Drivers and the Microsoft ODBC drivers for
> Oracle, and I get the same results.
> Some of the data looks just fine but some of the characters get replaced by
> ASCII values.
> I have installed RS SP1.
> Don't know what I am doing wrong...
8.1.7 oracle client is installed on RS Sever, and the result set is returning
strange data. I'm using a simple view that was created in the Oracle
database, that returns a small record set about 100 records. I'm seeing ASCII
values in place of some data, in the Data tab and the Preview tab and the
Published report.
I have tried using both Oracle Drivers and the Microsoft ODBC drivers for
Oracle, and I get the same results.
Some of the data looks just fine but some of the characters get replaced by
ASCII values.
I have installed RS SP1.
Don't know what I am doing wrong...Hi Gary
We are running SQL Reporting Services (SP1) against an Oracle 9i database
without any problems. In our case 9.2.0.4 client is used to talk with
9.2.0.4 database.
I suspect the problem you are having is related to version of Oracle client
you are using. If you have access to a client 9.2.0.4 or higher you may want
to give that a try.
"garyc" wrote:
> I have a RS report that is connecting to an Oracle Database (8.1.7), the
> 8.1.7 oracle client is installed on RS Sever, and the result set is returning
> strange data. I'm using a simple view that was created in the Oracle
> database, that returns a small record set about 100 records. I'm seeing ASCII
> values in place of some data, in the Data tab and the Preview tab and the
> Published report.
> I have tried using both Oracle Drivers and the Microsoft ODBC drivers for
> Oracle, and I get the same results.
> Some of the data looks just fine but some of the characters get replaced by
> ASCII values.
> I have installed RS SP1.
> Don't know what I am doing wrong...
Subscribe to:
Posts (Atom)