Order mismatch in selected data using Select query

Hello Experts,
We are upgrading from 4.6C to ECC6. I came across a select query on a custom table, which gives different order of data from 4.6C to ECC6. Data (no. of records) is same in both the servers but order is different. Table doesn't have indexes in either 4.6C or ECC6. What would be the cause for order change?
Further, in 4,6C, ECC6 and data bases order of the data is totally different. Select query have some conditions in where clause with select options, but in both the cases (4.6C and ECC6) I am not passing any data from the selection screen so am getting entire data from the data base.
Thanks in advance
Phalani

From the SAP service market place I found that
1. Unless we use SORT or ORDER BY clause the order of the data is not in our control.
2. In 4.6C, two same select queries never give the result in the same order, if the primary key contains character fields.
Thanks for your help.
Phalani

Similar Messages

IIf condition between dates using mdx query

Hi,
how to check IIF condition between dates using mdx query.
I able to check single year,plese check below mdx query.I need to check members between years.
my requirement is member is belongs to between years(2007 to 2010),display "yes" else "NO";
Could you please give me exact mdx query.
From,to-2007,2010,if member belongs to 2007to 2010 then disply "yes"else "no".
how to pass two members in IIf condition.
WITH
MEMBER Measures.[test]
AS Iif([Date].[Calendar Year].currentmember
is [Date].[Calendar Year].&[2007],"no","yes")
SELECT {Measures.[test]}
on 0
,[Product].[Subcategory].[Subcategory].MEMBERS * [Date].[Calendar Year].[Calendar Year]
ON 1
FROM [Adventure Works]
indu

Hi Sriindu,
consider the following:
WITH
MEMBER measures.[test] AS
IIF
Exists
[Date].[Calendar Year].CurrentMember
[Date].[Calendar Year].&[2007] : [Date].[Calendar Year].&[20010]
).Item(0)
IS
[Date].[Calendar Year].CurrentMember
,"yes"
,"no"
SELECT
{measures.[test]} ON 0
[Product].[Subcategory].[Subcategory].MEMBERS
[Date].[Calendar Year].[Calendar Year] ON 1
FROM [Adventure Works];
Philip,

Order mismatch in selected data from 4.6C to ECC6

Hello Experts,
We are upgrading from 4.6C to ECC6. I came across a select query on a custom table, which gives different order of data from 4.6C to ECC6. Data (no. of records) is same in both the servers but order is different. Table doesn't have indexes in either 4.6C or ECC6. What would be the cause for order change?
Further, in 4,6C, ECC6 and data bases order of the data is totally different. Select query have some conditions in where clause with select options, but in both the cases (4.6C and ECC6) I am not passing any data from the selection screen so am getting entire data from the data base.
Thanks in advance
Phalani

This kind of behaviour is usually encountered when "kernel" or "database" are upgraded. The Open-SQL optimizer may choice another way to select the data that change the order they are provided to the programs. Correct programs sort the data they receive or use explicit ORDER BY clause in their SELECT statement.
Correct the report(s) by adding the 'ORDER BY' clause to the SELECT statements. If the clause is not specified then the records may be returned in a random sequence. Even if the records are usually returned in primary key sequence, you should not rely on this happening.
Perform some [Search for SAP Notes|http://service.sap.com/notes] on "primary key" keyword for your systems.
Regards,
Raymond

Order by in select query

Hi guys
i am getting problem in my select query which is creating one long string, it is working fine without order by but when i run this with order by it get hanged. this query using on function to concatenate multiple strings in one string as well
Please help me what i am doing wroing.
select
regexp_substr(pe.emp_id,'[0-9]+$') ||'|'||
p.title ||'|'||
cdt.text||'|'||
j.tit||staff_pcb.host_pub_tit ||'|'||
jou.vol ||pcb.vol||pcp.vol ||'|'||
jou.journal_number||'|'||
substr (jou.pages||pcb.pages||pcp.pages, 1, instr (jou.pages||pcb.pages||pcp.pages, '-')-1)||'|'||
substr (jou.pages||pcb.pages||pcp.pages, instr (jou.pages||pcb.pages||pcp.pages, '-')+1)||'|'||
pcb.st||jou.st||pcp.st||'|'||
concat_aut(p.id)
from pub p,journal j,caa c,person pe,class_t cdt,
(select pcj.pub_id,pcj.vol,pcj.pages,pcja.st
From pub_C_j_a pcja,pc_j pcj
where pcja.pub_c_j_id = pcj.pub_id)jou,
(select pcb.pub_id,pcb.host_pub_tit,pcb.pages,pcb.vol,pcba.st
From pub_c_b pcb,pub_c_b_a pcba
where pcba.pub_c_b_id = pcb.pub_id)pcb,
(select pcp.pub_id,pcp.vol,pcp.pages,pcpa.st
From pub_c_p pcp,pub_c_p_a pcpa
where pcpa.pub_c_p_id = pcp.pub_id)pcp,
where
p.id = c.pub_id
and c.person_id = pe.id
and p.t_c_id = cdt.c_id
and p.journal_id = j.id(+)
and p.id = jou.publication_id(+)
and p.id = pcb.publication_id(+)
and p.id = pcp.publication_id(+)
and substr(regexp_substr(pe.emp_id,'[0-9]+$'),1,2) in('77','78')
and pcb.st||jou.st||pcp.st='1' order by pe.emp_id
;

Not sure about why it's only hanging on order by but looks like your query causing full table scan because of this in WHERE clause:
and pcb.st||jou.st||pcp.st='1'
I would change it to something like:
and ((pcb.st='1' and jou.st = '' and pcp.st = '') or (jou.st='1' and pcb.st = '' and pcp.st = '') or (pcp.st='1' and jou.st = '' and pcb.st = '') )

Failing to import data using SQL query

I am trying to import data from a sql query to analysis service in SSDT. i went to model > import from data source >Microsoft SQL Server and i chose from query. I typed my query which works fin in SSMS and when i press import if gives me a error. what
is going on please help here is the error
DirectQuery error: All tables used when querying in DirectQuery Mode must be from a single relational Data Source.
here is my query:
USE AdventureWorks2012;
SELECT SalesPersonID, [29484] Brian, [29485] Peter, [29486] Frank, [29518] Sarah, [29519] Janet, [29520] Alice, [29608] Martin, [29609] Patrick, [29610] Wasu, [29780] Samanyika, [29781] Vladmire
FROM
SELECT SalesOrderID, SalesPersonID, CustomerID
FROM Sales.SalesOrderHeader
WHERE SalesPersonID IS NOT NULL
As P PIVOT
COUNT (SalesOrderID) FOR CustomerID IN ([29484], [29485], [29486], [29518], [29519], [29520], [29608], [29609], [29610], [29780], [29781])
AS PVT
ORDER BY SalesPersonID ASC;

Hi,
When importing data from relational data source, the steps are:
In SQL Server Data Tools (SSDT), click the Model menu, and then click Import
from Data Source.
On the Connect to a Data Source page, select the type of database to connect to, and then click Next.
Follow the steps in the Table Import Wizard. On subsequent pages, you will be able to select specific tables and views or apply filters by using the Select
Tables and Views page or by creating a SQL query on Specify a SQL Query page.
As we can see on the steps, we need to select one database, so we needn't to write "USE AdventureWorks2012" on the query.
Reference:Import from a Relational Data Source
Regards,
Charlie Liao
TechNet Community Support

Default in "order by" in select query may be different in 11g from 10g

Hi, I am wondering if when you are querying a table in 11g and you use the order by clause and there is more than one occurrences with the same values in the order by, if the 11g default is different than from 10g. For instance.
DECLARE MHBulk CURSOR FOR
select invoice_nbr,
               customer_nbr,
                post_century,
                post_yymmdd,
     from CUSTOMERS
     where customer_nbr = 1234
     order by
           post_century,
           post_yymmdd;
If you have more than one occurrence of the same customer_nbr, post_century, and post_yymmdd it looks like the default order for how 11g retrieves the records is a bit different than the 10g default.   Does anyone know why this is?

JustinCave wrote:
In any version of Oracle, if your ORDER BY does not specify a completely unique order, the order in which ties are broken is arbitrary. Realistically, it will depend on things like the specific query plan chosen, the physical order of data on disk, which parallel query slave read the row, etc. So it is subject to change within the same version of the database when, for example, a query plan changes. It is entirely plausible that the results would be returned in the same order for years and then suddenly change one day. If you care about the order of ties, you should add additional logic to the ORDER BY clause that specifies the order you want rather than assuming that there is a "default" order applied.
Upgrading to a new version of Oracle, particularly if that was done via something like exporting and importing the data so that you're changing the Oracle version and the physical order of data and various object statistics, it is not at all unusual that you would get slightly different plans and that ties would be broken differently. I wouldn't assume that there is a new "default" order, though. It's entirely possible for the order to change if the query plan changes again.
Justin
Well, what happened is we've gotten a few new customers and those customers are on Linux oracle 11g. All the other customers are on regular unix 10g. So, it sounds like we've just been lucky with the order until now. In our system we have a sequence number column which is unique and I solved this by adding that in the ORDER BY clause. I was just curious as to whether oracle had a default when it found duplicate fields in the ORDER BY clause. I can't imagine that oracle doesn't have a default like unique_session_id or something like that.

Column order in a select * query

Suppose I have 256 columns in a table and if I query select * from tablename ,what will the column order in the result.Will this be always same as order as in created statement?

If the columns were all in the original CREATE TABLE statement, then, yes, the ordering of columns in SELECT * FROM will match the order from the CREATE TABLE statement.
If columns were added after the initial create, then the original columns will come first (in their order) and then the added columns (in their order), etc.
In the real world, it is a very bad practice to use SELECT * and have expectations about the ordering of the columns. Consider this scenario:
Table T was created and implemented in production a year ago with columns A, B, C
A project started up three months ago and added column D but the project is stalled in development. So in development T had columns A, B, C, D
Later, another project started up and added columns E and F. Unlike the other project, it has progressed to Test. and Production. So in Development the table has columns A, B, C, D, E, F. In test and production the table has columns A, B, C, E, F.
Now that stuck project has progressed to Test and Production. Now T has columns A, B, C, E, F, D in test and production but still A, B, C, D, E, F in development.
Until someone notices and decides to fix it (where and how?)
Be very, very careful about using SELECT *.

Filtering data using sql query

I am pulling data from a database which is a log of events type data. It consists of entries at a certain time, entry text, and even has an entry type. One of those types is known as "open item." This entry has the potential to be present on multiple days over and over. The problem is I only want the "open item" entries to appear that occur on the most recent day of a date range that is input by the user, but also not excluding all other entries from all of the other days that are not "open item." The "is_open=0" is what is excluding the "open items" from all of the entries, but now there needs to be a way to include (ie. "is_open=1") the "open items" on the current day. My query is as follows:
SELECT     b.log_title, a.log_id, MIN(a.shift_date) AS shift_date,a.shift_id,a.operator_position_descr,a.is_open,a.user_id, MAX(a.annotation_number) AS annotation_number,
                      a.entry_text, MIN(a.entry_time) AS entry_time,a.font_num
FROM         log_entries AS a INNER JOIN
                      log_list AS b ON a.log_id = b.log_id
WHERE ((entry_type_code NOT LIKE '%NLOG%'
            AND entry_type_code <> 'daily'
            AND entry_type_code <> 'shift act') or entry_type_code is null)
          AND (is_open = 0 AND is_standing_order = 0 AND is_rolled_forward is null)
GROUP BY b.log_title, a.log_id, a.user_id,a.shift_id,a.operator_position_descr,a.is_open, a.entry_text,a.font_num
Order by entry_time

Change you WHERE clause to something like this... (You didn't say what kind of DB you were using so I used SQL Server syntax)
WHERE ((entry_type_code NOT LIKE '%NLOG%'
AND entry_type_code IN ('daily', 'shift act')) OR entry_type_code is null)
AND (is_open = 0 OR (is_open = 1 AND a.shift_date >= DateAdd(dd, DateDiff(dd, 0, GetDate()), 0)))
AND is_standing_order = 0
AND is_rolled_forward is null
The biggest change is this... AND (is_open = 0 OR (is_open = 1 AND a.shift_date >= DateAdd(dd, DateDiff(dd, 0, GetDate()), 0))).
It allows is_open = 1 if the shift_date is => midnight of the of the current day.
HTH,
Jason

Help needed in inserting data using a query

I need to have some data as a result of the following query:
select sched_num,load_id,ord_id,split_id from ord_load_seq
where (ord_id,split_id,sched_num) in (select ord_id,split_id,sched_num from ord_load_seq where seq_num = '2'
group by ord_id,split_id,sched_num having count(1) >1)
order by ord_id,split_id,sched_num;
But currently it retunrns no rows. The problem is in the having count(1)> 1 clause.
When i make =1, it returns rows. But no rows on > 1. I even tried inserting some rows to get the result >1
But still the query on a whole returns no rows.Please help.

ohhh... lets start our lesson children:
here is code to consider:
Connected to Oracle Database 10g Enterprise Edition Release 10.2.0.2.0
SQL>
SQL> --Q1
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 3 fld3 from dual
7 union all
8 select 1 fld1, 2 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 3 fld2, 3 fld3 from dual)
11 select fld1, fld2, fld3 from tbl
12 order by fld1, fld2, fld3
13 /
      FLD1       FLD2       FLD3
         1          1          1
         1          1          2
         1          1          3
         1          2          3
         1          3          3
SQL> --Q2
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 3 fld3 from dual
7 union all
8 select 1 fld1, 2 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 3 fld2, 3 fld3 from dual)
11 select fld1, fld2, fld3, count(1) from tbl
12 group by fld1,fld2,fld3
13 order by fld1, fld2, fld3
14 /
      FLD1       FLD2       FLD3   COUNT(1)
         1          1          1          1
         1          1          2          1
         1          1          3          1
         1          2          3          1
         1          3          3          1
SQL> --Q3
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 3 fld3 from dual
7 union all
8 select 1 fld1, 2 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 3 fld2, 3 fld3 from dual)
11 select fld1, fld2, fld3 from tbl
12 group by fld1,fld2,fld3
13 having count(1) > 1
14 order by fld1, fld2, fld3
15 /
      FLD1       FLD2       FLD3
SQL> --Q4
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 2 fld3 from dual
7 union all
8 select 1 fld1, 1 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 2 fld2, 3 fld3 from dual
11 union all
12 select 1 fld1, 3 fld2, 3 fld3 from dual)
13 select fld1, fld2, fld3 from tbl
14 order by fld1, fld2, fld3
15 /
      FLD1       FLD2       FLD3
         1          1          1
         1          1          2
         1          1          2
         1          1          3
         1          2          3
         1          3          3
6 rows selected
SQL> --Q5
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 2 fld3 from dual -- inserted duplicate row
7 union all
8 select 1 fld1, 1 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 2 fld2, 3 fld3 from dual
11 union all
12 select 1 fld1, 3 fld2, 3 fld3 from dual)
13 select fld1, fld2, fld3, count(1)   from tbl
14 group by fld1,fld2,fld3
15 order by fld1, fld2, fld3
16 /
      FLD1       FLD2       FLD3   COUNT(1)
         1          1          1          1
         1          1          2          2
         1          1          3          1
         1          2          3          1
         1          3          3          1
SQL> --Q6
SQL> with tbl as
2 (select 1 fld1, 1 fld2, 1 fld3 from dual
3 union all
4 select 1 fld1, 1 fld2, 2 fld3 from dual
5 union all
6 select 1 fld1, 1 fld2, 2 fld3 from dual -- inserted duplicate row
7 union all
8 select 1 fld1, 1 fld2, 3 fld3 from dual
9 union all
10 select 1 fld1, 2 fld2, 3 fld3 from dual
11 union all
12 select 1 fld1, 3 fld2, 3 fld3 from dual)
13 select fld1, fld2, fld3 from tbl
14 group by fld1,fld2,fld3
15 having count(1) > 1
16 order by fld1, fld2, fld3
17 /
      FLD1       FLD2       FLD3
         1          1          2
SQL> Q1. As you may see we have an bunch of data where each row is unique combination of the columns value.
Q2. lets try to group it by fld1 and fld2 and fld3 columns. We don't expect any miracle and got the same data as Q1. Why? Because each row is unique combination(group) in scope of fld1, fld2, fld3 - count(1) exactly shows us that.
Q3. Q2 is explanation why we got no rows filter (having count(1) > 1)because the result of the clause is always false.
Q4. Lets put some duplication.
Q5. Now we got some new result. Count shows us group with more then one rows.
Q6. Shows us exact group which we've found in Q5.

How to get AD reconciliation event data using SQL query.

Hi All,
I was trying to get all recon data for Target AD Recon from OIM db, but it seems that in OIM 11g we do not have recon data in tables: RCU, RCE, etc. We have to use "RECON_..." tables, my question is does any bady has sample query to fetch data from recon tables?
best
mp

It should be a question of linking your AD user reconciliation table (beginning RA_) with the RECON_EVENTS table by RE_KEY.
In my case I have OID user reconciliation (so my RA_ table will be different to yours) and can use something like:
SELECT RECON_EVENTS.RE_STATUS, RA_OIDUSER5.* FROM RECON_EVENTS, RA_OIDUSER5 WHERE RECON_EVENTS.RE_KEY = RA_OIDUSER5.RE_KEY;
Obviously it then depends on what information you want to filter on and retrieve.

How to display contineous 10 dates from today date using SQL query

Hi All
I need a requirement of displaying 10 dates and numbers contineously, I have written a query for getting 10 numbers like this
select rownum from dual where rownum < 10 connect by rownum = level;
But i am not able to get the dates, Can any one help me ?

Hi Satyaki,
No issues on my 9.2.0.8 client.
SQL*Plus: Release 9.2.0.8.0 - Production on Sat May 3 09:49:17 2008
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
SQL> conn test/test@db10g
Connected.
SQL> select rownum, trunc(sysdate) + rownum from dual where rownum <= 10 connect by rownum = level;
    ROWNUM TRUNC(SYS
         1 04-MAY-08
         2 05-MAY-08
         3 06-MAY-08
         4 07-MAY-08
         5 08-MAY-08
         6 09-MAY-08
         7 10-MAY-08
         8 11-MAY-08
         9 12-MAY-08
        10 13-MAY-08
10 rows selected.
SQL> Regards

Swap table data using Update query

Hi,
If I have a table T with 4 rows & 2 columns A & B. A has
values 1,2,3,4. and B has 10,20,30,40.
A  B
1  10
2  20
3  30
4  40
Write an Update SQL query which will Swap the values of A & B for all records.
(Do not use a sub-query)
I faced it as a Multiple choice question. I don't remember the choices exactly but all choices were single line Update queries.
Thanks in advance.

Maybe I'm missing something but this is what I came up with:
SQL> CREATE TABLE TEST_TAB(A number, B number);
Table created.
SQL> INSERT INTO TEST_TAB VALUES(1,10);
1 row created.
SQL> INSERT INTO TEST_TAB VALUES(2,20);
1 row created.
SQL> INSERT INTO TEST_TAB VALUES(3,30);
1 row created.
SQL> INSERT INTO TEST_TAB VALUES(4,40);
1 row created.
SQL> SELECT * FROM test_tab;
         A          B
         1         10
         2         20
         3         30
         4         40
SQL> UPDATE test_tab SET A=B, B=A;
4 rows updated.
SQL> SELECT * FROM test_tab;
         A          B
        10          1
        20          2
        30          3
        40          4
SQL>Hope this helps!

Master Data Deletion - Data used in query

Hi all,
I work with BI 7. I noticed a strange thing. I wanted to delete data from a master data object, but the operation is not allowed. If I see the logs for "where used list", I see a query. I deleted data from DSO objects and infocubes, and I thought it was enough.
Any idea?
Thanks a lot.

Hello Gianluca,
To be able to delete master data, no transaction data must exist for this master data, the master data must no longer be used as attribute value, hierarchies to this master data must not exist, and it must not be used in a query.
If you look at a record in the SID table, you can check the usage flag DATAFL. If the value is X, it is used or has been used by another BW oject.
But to answer your question, do delete the master data, it must no longer be used in a query.
Best Regards,
Vincent

Doubt in Select Query

Hi,
I want to get data from 2 table which query is effeicient one
by seeing performance...
1. using for all entries statement
2. Get data using join query.
Thanks in advance

FOR ALL ENTRIES is better for some reasons,
--> If the data is huge , then some times JOIN may end up in TIME OUT error
--> FOR ALL ENTRIES will use the INDEXEs of the tables and improves the performance

Using Bex query through Query Browser - add result set to Cell range?

Hi all,
I have up to now used Xcelsius on NW with BICS connection towards BEx queries.
When I use this connection via Data Manager in Xcelsius, I'm able to use the Cross-table functionality to insert my entire dataset into a cell Range.
But now I have an requirement to save my Dashboards on the BO server, so I need to go through Query Browser to extract data from SAP BW to my Dashboard.
When inserting the query in the Query browser and I want to assign the data to Excel ranges, it seems like I need to do this for each individual Infoobject and Key Figure. Since I have many Key Figures (30-50), this is a lenghty process.
Is there a way to insert the entire dataset from Query into a cell range (both rows and columns), or do I need to do this for each Key Figure as indicated in the screen shot above?
--> I know that the Query Browser flattens my cross-table from the Query, but still, it should be possible to avoid to input all my Key Figures (30-50) individually to corresponding cells in my Excel part of the Dashboard.
regards
Oddmar Lid
asdfasf

Hi Oddmar,
If you want to display the data retrieved from the query you could map the data using the Query data.
Pull scorecard component (for example) and in properties if you see the Display data click on the dropdown and select Query data.
Hope this helps !

Order mismatch in selected data using Select query

Similar Messages

Maybe you are looking for