1......100 from dual

Hi,
How to print 1..100 using dual table in a SQL statement(no pl/sql) ?
Regards
gopureddi

I should have added this works only in 10g :
SQL> select level from dual connect by level <= 100;
LEVEL
1
2
3
4
5
6
7
8
9
10
11
in 9i :
SQL> select level from dual connect by level <= 100;
LEVEL
1

Similar Messages

How to generate multiple records on a single sql from dual table

I wanted to generate ten sequence nos in a single sql statement from dual table.
Is there any way to use that.
I think somebody can help me on this by using level clause

I'm not 100% sure if I understand your requirement: Do you really want to use an Oracle Sequence, as Alex already demonstrated?
Or just a 'one-time-bunch-of-sequential-numbers'.
In the latter case you can just select level:
SQL> select level
2 from   dual
3 connect by level <= 10;
     LEVEL
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
10 rows selected.

Select sysdate from dual, why?

I have been looking at the SQL that is generated during fetches, updates
etc. Why is the SELECT SYSDATE FROM DUAL being executed prior to every sql
statement being executed?
If I had to guess I would say it has to do with locking or logging but
don't really have a clue!
I suppose the final question is ... how can I switch it off!
Thanks in advance
Matt

Have not seen oracle just dropping connection. In any case if you are using
oracle JDBC 2 it has error event which lets you handle error conditions plus
PooledConnection.getConnection() raises exception if its underlying physical
connection is dead so you can catch it and discard old PooledConnection (and
do not return it to the pool) and create new one without error ever going to
the user. No need for "select sysdate from dual" once in a while or god
forbid on every getConnection() /* hope oracle handles it more efficiently
*/ I stress tested my pool implementation - running many threads on JDO
reads and killing connections on server left and right and Kodo never saw
killed connections - it all was handled within pool itself without anything
like "select sysdate from dual". What I am trying to say is at least as far
as Oracle JDBC concerned their PooledConnection takes care of this
"Patrick Linskey" <[email protected]> wrote in message
news:[email protected]...
Matt,
We issue this to validate that connections retrieved from the pool are,
in fact, open. Oracle has a nasty habit of dropping connections (you'd
be surprised with some of the ways that Oracle sucks...), so it tends
to be good to do this check relatively frequently.
The frequency at which it occurs depends on a number of factors. First,
it depends on whether or not you are using a connection pool. I believe
that if you do not use the conn pool, then we probably check every time
you get a connection. Second, it depends on the setting of the
com.solarmetric.kodo.impl.jdbc.ConnectionTestTimeout setting, which
specifies how often to check connections for validity. I believe that
it's a number in seconds, but I'm not 100% sure; the configuration
section of our docs will have the right answer. Third, it depends on
the setting of com.solarmetric.kodo.impl.jdbc.ConnectionRetainMode,
which controls when we get connections and when we release them.
Finally, you can turn it off altogether by setting the
com.solarmetric.kodo.impl.jdbc.DBDictionaryProperties property to
contain the string 'ValidateConnections=false', which will turn off
connection validation altogether.
-Patrick
Patrick Linskey
SolarMetric Inc.

Strange "select level from dual connect by level :N" behavior

Oracle SQL Deveoper version:1.1.0.23
Oracle database server version:
select * from v$version;
Oracle9i Enterprise Edition Release 9.2.0.7.0 - Production
PL/SQL Release 9.2.0.7.0 - Production
"CORE 9.2.0.7.0 Production"
TNS for 32-bit Windows: Version 9.2.0.7.0 - Production
NLSRTL Version 9.2.0.7.0 - Production
Run the following queries in sequence as a script and look at number of rows returned:
select 1, level l from dual connect by level < 1;
select 2, level l from dual connect by level < 2;
select 3, level l from dual connect by level < 3;
select 4, level l from dual connect by level < 4;
select 5, level l from dual connect by level < 5;
select 6, level l from dual connect by level < 6;
select 7, level l from dual connect by level < 7;
select 8, level l from dual connect by level < 8;
select 9, level l from dual connect by level < 9;
select 10, level l from dual connect by level < 10;
select 11, level l from dual connect by level < 11;
select 12, level l from dual connect by level < 12;
1 L
1 1
1 rows selected
2 L
2 1
1 rows selected
3 L
3 1
3 2
2 rows selected
4 L
4 1
4 2
4 3
3 rows selected
5 L
5 1
5 2
5 3
5 4
4 rows selected
6 L
6 1
6 2
6 3
6 4
6 5
5 rows selected
7 L
7 1
7 2
7 3
7 4
7 5
7 6
6 rows selected
8 L
8 1
8 2
8 3
8 4
8 5
8 6
8 7
7 rows selected
9 L
9 1
9 2
9 3
9 4
9 5
9 6
9 7
9 8
8 rows selected
10 L
10 1
10 2
10 3
10 4
10 5
10 6
10 7
10 8
10 9
9 rows selected
11 L
11 1
11 2
11 3
11 4
11 5
11 6
11 7
11 8
11 9
11 10
10 rows selected
12 L
12 1
12 2
12 3
12 4
12 5
12 6
12 7
12 8
12 9
12 10
10 rows selected
1000 L
1000 1
1000 2
1000 3
1000 4
1000 5
1000 6
1000 7
1000 8
1000 9
1000 10
10 rows selected
-- wrong result. return 1 row, should return 0 row
-- May have some automagical code saw dual and return 1 row.
select 1, level l from dual connect by level < 1;
-- wrong result. Should return 11 rows, but it stopped at 10 rows.
select 12, level l from dual connect by level < 12;
select 1000, level from dual connect by level < 1000;

""By default, when Oracle JDBC executes a query, it receives the result set 10 rows at a time from the database cursor.""
Sql dev overrides the default and fetches 50 (1.0) or 100 (1.1) rows at a time. In 1.1 this is configurable.
"I don't HATE any SQL, as long as it's valid."
That statement is absurd. So, if someone tells you a way to write your query that it allows it to run faster, you wouldn't care? You should hate some SQL, some of it is pure crap and needs to be rewritten. This is what we call 'junk code'. As a developer, your goal should be to write maintainable code. In short, if you have the choice between something that looks fancy and something that is legible and maintainable, choose the maintainability. Sorry, but if I ever saw that statement in a code review I would have made you rewrite it. Also, anything that runs differently on different databases should be avoided, its that simple.

Mutating Problem when insert while selecting data from dual

Hi All,
we have a table
test (
ID NUMBER
NAME VARCHAR2(100)) and a before insert trigger
create or replace trigger test1
before insert on test
for each row
begin
select decode(max(id),null,1,max(id)+1) into :new.id from test;
end;
i am able to insert values by using
"insert into test(name) values('test1')" with out any issues.
when i am inserting the values
"insert into test(name) select 'test1' from dual" i am getting error
ORA-04091: table SCOTT.TEST is mutating, trigger/function may not see it
Could someone please advice why i am getting error in second scenario.
Thanks in Advance
Prasad

Prasad
try
insert into test (name) values ((select 'test1' from dual));
Frank

Parallel invokation of illustrator from dual core machine.

We have been using CS3 for long time. The demand of the utility (which uses illustrator) has been increased a lot. As result we have run multiple illustrators at the same time (parallel) from dual core Windows 2033 server. We are unable to find a suitable example to do this task. It will be really helpful if anyone can provide help in this regard.
Similarly currently illustrator popping-up illustrator while creating the image. Is there any option to stop popping-up illustrator on the screen while creating the image. I know we have a command for adobe normalize while converting post script file to PDF.
Thanks in advance.
Regards,
Bhasya.

Adobe desktop products are unsupported on server operating systems, so you are 100% on your own on this one, even more so for a legacy product liek CS3. That aside, there is no way to run multiple instances of AI at the same tiem other than through virtualization, but even then you still have to deal with the technical and legal aspects of licensing....
Mylenium

Yoga 3 Pro or Yoga 2 Pro? Opinions from dual owners please

Hi,
I'm tossing up between the Yoga 2 Pro and the Yoga 3 Pro.
I like the design of the 3 very much, and the weight.
I will run Ofiice apps, Photoshop CS6 on the road, and lightroom 5.7.
Any pros and cons from dual users with experience of this software in daily use would be apprecited.
Regards,
David.

I will be doing a video review on my experience soon but here are some points I can tell you already -
1) It is a great travelling companion, movies, videos, songs, small gaming is all perfect.
2) Battery life is really nice, I watched 3.5 hour long Lord of the rings and still had about 60% left
3) Pretty light and rotates 360 degrees, very comfortable to use in any and every mode.
4) Does not get hot, remains cool.
5) Exquisite looks especially the Bright orange color, which is my fav.
6) Screen is beautiful and everything is crisp (I use it at 1080p)
7) Some apps are very useful like ShareIt, transfer data from android, iphone etc in seconds over a wifi network.
8) Performance is good for browsing as well.
9) Boots in 3-5 seconds
10) Keyoard is backlit and very nice but shallow, trackpad is an absolute delight.
2 finger taps, 2 finger swipes, 3 finger swipes, 2 finger zoom, 2 finger scrolling of pages, everything is smooth and responsive.
11) Sound quality is awesome as well, very loud and clear. Enough bass to provide enjoyable experience.
What you should not expect -
1) Heavy graphics intensive tasks.
2) Lags a bit on some web pages scrolling (NOT noticeable)
3) Some software which are useless and can be uninstalled.
I typed this whole post on my Yoga 3 Pro and it is a great for me to type on this beauty.
Going to watch a movie now.
Current System - Lenovo Y510P, GT755M SLI, 8GB RAM, 1TB HDD + 24GB SSD
When your plans fail, your Real Story begins!!

Display a column with a list of values from DUAL;

Hello, how can I display something like:
COL1
A
B
C
D
E
F

On my planet these are 'words' ;) :
SQL> select dbms_random.string('a', level)
2 from dual
3 connect by level <= 10;
DBMS_RANDOM.STRING('A',LEVEL)
r
BB
pmE
bHri
aZZDC
FzOZBf
fXfPyIP
LOqrQbnK
AYzNgHcGc
lvxRvflHXj
10 rows selected.
SQL> select dbms_random.string('a', dbms_random.value(10, 50))
2 from dual
3 connect by level <= 10;
DBMS_RANDOM.STRING('A',DBMS_RANDOM.VALUE(10,50))
AKExSSzIJOkabCT
VCHHRTcgtnBSlZyRmsITqzWqVNgDnPdnxmt
qLnuyGQErxZvHttRBpWrPoLEiGUra
rcllLkyZXiwvlcGaUoEAeMbEdZSHlZBIgdilgGTQzxlIUW
ftnGzDCzUBEsxKzhDLfgS
aqUmpIzjqkWMZGpbAchhbiqFBUKdRaDwFAgdP
nskqNdzmlAKlsGOJpbKfQHiCfvkDpoFcTkTvKaUcGLFDkgjsX
fwkxBSoVuDSEOFldgDhilvm
iAnSvDMQzamGvIFNydfjhr
lHRPUnKgOxYQwWAltdELJNqqfgMRSmbBpSDnjOsfx
10 rows selected.http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14258/d_random.htm#CDEJAHCB

Select from dual into a variable through db links

HI,I need to select value of current_timestamp in to a variable d1 of a remote database inside function.
function (db_lnk varchar2)return number
as
dbl varchar2(10);
begin
dbl:=db_lnk;
select current_timestamp into d1 from dual@dbl;
end;
but getting error table or v iew does not exist.
if i do
select current_timestamp into d1 from dual@'||dbl||';
then it says database link name expected.
How to achieve this?
Thanks

Peter Gjelstrup wrote:
Foreign languages, foreign languages :-){noformat}*grins*{noformat} I know - and your English is miles better than my Danish (I'm assuming - hopefully correctly?! - that that's your 1st language based on your location!) which I don't even know any words of! *{:-)

JDBC Lookup in PI 7.1 - SELECT ? FROM DUAL and Connection timed out

Hi,
We have a scenarios (Idoc to JMS) with JDBC lookup. We have used graphical JDBC lookup functionality.
We are reading country names for a given country code from SAP in an external database table. The query is so simple. That should not take much time
Now the actual issues is,
When we are executing the scenario, its taking quite a long time. Almost 6 minutes to excute a mapping. Which causing high performance issue in the Porduction.
We started the inviestigation about found some interesting stuff. Here we have used Willy Introscope for the investigation.
1. First few messages are taking quite a long time. LIke 6 minutes per messages. As i can see in the log i am getting below error in Willy,
I dont know why PI is executing below queury apart from real secelt query to fetch the country name. I am getting below error: Error Message: Backends|ABCD2 mydatabase01-1526 (Oracle DB)|SQL|Dynamic|Query|SELECT ? FROM DUAL: java.sql.SQLException: Io exception: Connection timed out
2. After couple of messages, interface works very normal. I mean rest of the messages works pretty fine.
Please let me know if you have any idea about this error. What could be the problem for the issue.
Thank you in advnace.
Best Regards,
Prasad.

Did you check how many SQL requests were executed per one message ? Do you have a log of these SQL requests ?
I assume that the country table is quite small, so that lookup should not be an issue.
About this:
>Message: Backends|ABCD2 mydatabase01-1526 (Oracle DB)|SQL|Dynamic|Query|SELECT ? FROM DUAL: >java.sql.SQLException: Io exception: Connection timed out
1. I only know SELECT * FROM DUAL, not SELECT ? FROM DUAL. Better use the former
2. the exception means that the database server can not be reached => check your network configuration
So I assume that there is a network (performance) problem between PI and this Oracle server. Or the Oracle Server is so overloaded that it has (sometimes) problems in processing new requests.
CSY
Edited by: Christian Sy on Mar 9, 2010 10:17 AM

[11g] increasing efforts for select xmlserialize(content(...))) into ... from dual

I wonder, that i don't get any feedback here (please see below).
Can somebody first just confirm this observation?
Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production
PL/SQL Release 11.2.0.3.0 - Production
"CORE 11.2.0.3.0 Production"
TNS for Linux: Version 11.2.0.3.0 - Production
NLSRTL Version 11.2.0.3.0 - Production
NLSRTL 11.2.0.3.0 Production
Oracle Database 11g Enterprise Edition 11.2.0.3.0 64bit Production
PL/SQL 11.2.0.3.0 Production
TNS for Linux: 11.2.0.3.0 Production
I recognized for my application, that there is an increasing efforts (in terms of time) for executing
(in a pl/sql package procedure invoked by a BIU trigger)
SELECT XMLSERIALIZE(CONTENT(p_xml_data)) INTO v_xml_clob FROM DUAL;
Here is a little demo sqlplus script:
create or replace package test_pkg
as
       procedure check_xml(
         p_xml_data in out nocopy xmltype);
end;
create or replace package body test_pkg
as
       procedure log(
         p_txt in varchar2)
       is
         ts     TIMESTAMP WITH TIME ZONE := systimestamp;
       begin
         dbms_output.put_line(to_char(ts, 'hhmiss.FF3') || ':' || p_txt);
       end;
       procedure check_xml(
         p_xml_data in out nocopy xmltype)
       is
         v_xml_clob CLOB;
         v_len NUMBER;
       begin
         log('check_xml(): enter');
         SELECT XMLSERIALIZE(CONTENT(p_xml_data)) INTO v_xml_clob FROM DUAL;
         log(' serialized');
         v_len := dbms_lob.getlength(v_xml_clob);
         log('check_xml(): done - length = #' || v_len);
       exception
       when others then
           oerror_pkg.set_ora_error;
           raise;
       end;
end;
drop table test_table;
create table test_table(
       tid number(19,0),
       data xmltype
create or replace trigger BIU_TEST_TABLE
before insert or update on test_Table
for each row
declare
begin
   test_pkg.check_xml(:new.data);
end;
insert into test_table(tid, data)
select ctr_tab.ctr, '<root><node>' || ctr_tab.ctr || '</node></root>'
from ( SELECT LEVEL ctr
       FROM dual
       CONNECT BY LEVEL <= 200) ctr_tab;
The output is going like this
021543.204:check_xml(): enter
021543.204: serialized
021543.204:check_xml(): done - length = #32
021543.206:check_xml(): enter
021543.206: serialized
021543.206:check_xml(): done - length = #32
021543.207:check_xml(): enter
021543.208: serialized
021543.208:check_xml(): done - length = #32
021543.209:check_xml(): enter
021543.210: serialized
021543.210:check_xml(): done - length = #32
021543.211:check_xml(): enter
021543.212: serialized
021543.212:check_xml(): done - length = #32
021543.214:check_xml(): enter
021543.214: serialized
021543.214:check_xml(): done - length = #32
021549.625:check_xml(): enter
021549.664: serialized
021549.665:check_xml(): done - length = #34
021549.708:check_xml(): enter
021549.746: serialized
021549.747:check_xml(): done - length = #34
021549.791:check_xml(): enter
021549.829: serialized
021549.830:check_xml(): done - length = #34
021549.874:check_xml(): enter
021549.912: serialized
021549.913:check_xml(): done - length = #34
When i filter it with a little perl script to extract the efforts (xmlserialize / dbms_lob.getlength):
0 / 0
0 / 0
1 / 0
1 / 0
1 / 0
0 / 0
0 / 0
0 / 1
1 / 0
0 / 0
0 / 1
1 / 0
0 / 0
0 / 1
1 / 0
0 / 1
0 / 1
0 / 0
0 / 1
0 / 0
0 / 1
0 / 0
1 / 0
0 / 1
0 / 0
31 / 1
31 / 1
32 / 1
32 / 1
32 / 1
33 / 0
33 / 0
34 / 1
34 / 0
34 / 1
34 / 0
34 / 1
34 / 1
35 / 1
35 / 1
36 / 0
36 / 1
36 / 1
37 / 1
37 / 0
37 / 1
38 / 0
38 / 0
39 / 1
38 / 1
38 / 1
38 / 1
Unfortunately i can't easily change the way the insert from select is done (legacy code not under my control)
Can someone tell me, if there is a way starting with the trigger to avoid those increasing efforts?
- many thanks!
best regards,
Frank

[ Addendum - 24.01.2014:
This only worked for the test/demo program. It didn't work for my application.
   For my application i really had to move the logic to the INSERT before the trigger
   - similar to the 3rd posting for my monologue(!) here.
A little step further:
I also don't have the problem if i expand the code of the PROCEDURE test_pkg.check_xml(...) directly into the trigger
- instead of invoking the PROCEDURE in the trigger:
create or replace package test_pkg
as
       procedure log(
         p_txt in varchar2);
end;
create or replace package body test_pkg
as
       procedure log(
         p_txt in varchar2)
       is
         ts     TIMESTAMP WITH TIME ZONE := systimestamp;
       begin
         dbms_output.put_line(to_char(ts, 'hhmiss.FF3') || ':' || p_txt);
       end;
end;
drop table test_table;
create table test_table(
       tid number(19,0),
       data xmltype
create or replace trigger BIU_TEST_TABLE
before insert or update on test_Table
for each row
declare
begin
-- test_pkg.check_xml(:new.data);
       declare
         v_xml_clob CLOB;
         v_len NUMBER;
       begin
         test_pkg.log('check_xml(): enter');
         SELECT XMLSERIALIZE(CONTENT(:new.data)) INTO v_xml_clob FROM DUAL;
         test_pkg.log(' serialized');
         v_len := dbms_lob.getlength(v_xml_clob);
        test_pkg.log('check_xml(): done - length = #' || v_len);
       exception
       when others then
           oerror_pkg.set_ora_error;
           raise;
       end;
   end;
insert into test_table(tid, data)
select ctr_tab.ctr, '<root><node>' || ctr_tab.ctr || '</node></root>'
from ( SELECT LEVEL ctr
       FROM dual
       CONNECT BY LEVEL <= 200) ctr_tab;
-- rollback;
That gives some hint.
Does oracle have some problem / limit for invoking procedures (functions) from triggers?
Or only if those use certain features?
An issue about deterministic and re-entrance?
Well, that boxes me into a corner.
Because the single package procedure implements a functionality at a central place - i.e. a single central place to extend it or to fix it.
If i now have to expand its content (like a macro) into the triggers of the respective table i am in trouble.
Because those tables (and their triggers) are dynamically generated by a compiler tool of a c++ client.
This means for extension and fixes i need to change, test, deliver and deploy a list of c++ client processes :-(
Is there any way around?
How can i inform oracle that the invocation of a PL/SQL procedure is functionally identically with expanding the PL/SQL functions code into the trigger?
rgds,
Frank

Problem with: select 'c' as X from dual

Problem with 'select 'c' as X from dual'
I get 2 different results when I execute the above with SQLPlus (or java) depending on the instance I am connected to. For one instance the result is a single character and for the other the character is padded with blanks to 32 chars in the SQLPlus window (and java). Does anyone know what database setting causes this to happen? Is it a version issue ?
Test #1: Oracle 9.2.0.6 - SQLPlus result is padded with blanks
SQL*Plus: Release 9.2.0.1.0 - Production on Mon Dec 10 09:27:58 2007
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Connected to:
Oracle9i Enterprise Edition Release 9.2.0.6.0 - 64bit Production
With the Partitioning option
JServer Release 9.2.0.6.0 - Production
SQL> select 'c' as X from dual;
X
c
SQL>
Test #2 Oracle 9.2.0.1 SQLPlus result is a single character.
SQL*Plus: Release 9.2.0.1.0 - Production on Mon Dec 10 09:29:27 2007
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Connected to:
Oracle9i Enterprise Edition Release 9.2.0.1.0 - Production
With the Partitioning, OLAP and Oracle Data Mining options
JServer Release 9.2.0.1.0 - Production
SQL> select 'c' as X from dual;
X
c
SQL>

Using 9.2.0.6 On AIX 5.2 I get the single byte result:
UT1 > select 'c' as X from dual;
X
c
If the databases are on different Oracle Homes you may want to check the sqlplus global logon files for any set commands.
If you executed the two sql statements from different OS directories you may also want to check your sqlpath for sqlplus .logon files.
Try issueing clear columns and repeating the statement. Are the results the same?
HTH -- Mark D Powell --

Create a view with datatype from dual

I have Windows XP with 10g 10.2.0.1.0
I need to create a view from dual and with data type, such as NUMBER(5,2), NUMBER or VARCHAR2(20) .
such as:
create view test(view_test varchar2(20)) as (select view_test from dual)
Please help. Thanks in advance.

It does seem kind of odd, but you can do something like this with the CAST function:
SQL> create view dual_view as
2 select cast(null as number(5,2)) col1
3        ,cast(null as date)    col2
4        ,cast(null as varchar2(30)) col3
5 from   dual;
View created.
SQL> desc dual_view
Name                                      Null?    Type
COL1                                               NUMBER(5,2)
COL2                                               DATE
COL3                                               VARCHAR2(30)I used NULLs but you can use actual values if you want.

Delay due to selecting sysdate from dual

Our application accessing dummy table dual frequently for getting current date and time. But some times this could reduce our database performance.
How could we overcome this problem.
ASH report_
Top Events_
latch: library cache 29.87 "5015593920","214","0" 26.57 address number tries
Top SQL Statements_
0vnruqfxmtu6q Select to_number(to_char(Sysdate, 'YYYYMMDD')) from dual

orianmoon wrote:
Thanks for your replies,
Please check this below link which shows my database performance at the time of running a particular sql , which is selecting sysdate using table dual. It created Lbrary cache latch contention.
http://arun-itadmin.blogspot.com/2010/12/oracle-performance-screen.html
My question is
Is this the problem of frequent usage of table dual or usage of 'sysdate' ?
It there any way to get date and time other than using sysdate function?
Edited by: orianmoon on Dec 8, 2010 9:42 PMI don't see anything there that points the finger at SELECT SYSDATE FROM DUAL;

Bug: in ManagedDataAccess while fetch from "Select .., NULL, ... From dual"

Hi,
I have the problem with the new managed ODP.Net driver.
If I fetch data from a select which has column with a fixed value "NULL" the driver decides after a certain amount of data that the value is not null any more. Which causes a lot of problems especially if you try to hide not needed blob data.
The Problem somehow depends on the FetchSize of the command. It seems like the error occurs if more than one db round trip to fetch the result is necessary.
System: Windows 7 64 Bit
Platform: .net 4.0 x86
database: 11g Release 11.2.0.3.0 - 64bit Production
Oracle.ManagedDataAccess Version: 4.112.350
I created a small example to reproduce the problem.
Thanks for your help
Dominik
Stored Proc:
create or replace PROCEDURE TestNullField
v_IntPara IN NUMBER DEFAULT NULL ,
v_StrPara IN VARCHAR2 DEFAULT NULL,
cv_1 OUT SYS_REFCURSOR
AS
BEGIN
      OPEN cv_1 FOR
        select rownum, v_StrPara, NULL from dual connect by level <= v_IntPara;
        --select IDX, NULL, DESCRIPTION FROM TEST_BLOBTABLE;
END;C# Code:
using System;
using System.Text;
using Oracle.ManagedDataAccess.Client;
using System.Data;
namespace OracleBlobTest
    class Program
        private static string _connectionString = @"User ID=YourUser;Password=YourPwd;Data Source=YourServer:YourPort/YourSchema;";
        private static string _spName = @"TestNullField";
        private static string _strPara = @" Long test string";
        private static int _intPara = 200;
        static void Main(string[] args)
            using (OracleConnection connection = new OracleConnection(_connectionString))
                using (OracleCommand cmd = connection.CreateCommand())
                    cmd.CommandText = _spName;
                    cmd.CommandType = CommandType.StoredProcedure;
                    connection.Open();
                    string alongString = _strPara;
                    while (alongString.Length < 2000)
                        alongString += alongString;
                    alongString = alongString.Substring(0, 2000);
                    OracleCommandBuilder.DeriveParameters(cmd);
                    if (cmd.Parameters.Count > 0 && (cmd.Parameters[0]).Direction != ParameterDirection.ReturnValue)
                        cmd.Parameters[0].Value = _intPara;
                        cmd.Parameters[1].Value = alongString;
                    // change this to change the moment when it starts to go wrong
                    ///cmd.FetchSize = 5000;
                    using (OracleDataReader reader = cmd.ExecuteReader())
                        int count = 0;
                        while (reader.Read())
                            count++;
                            for (int idx = 0; idx < reader.FieldCount; idx++)
                                if (reader.GetName(idx) == "NULL")
                                    if (!reader.IsDBNull(idx))
                                        //something is very wrong here - why is not not null any more???
                                        Console.WriteLine("Fix NULL Field[{0}] {1} is not null >{2}< in row {3} ", idx, reader.GetName(idx), reader[idx], count);
                        Console.WriteLine("Rows found: " + count);
                connection.Close();
            Console.WriteLine("done press enter");
            Console.ReadLine();
}Edited by: user540519 on 10.12.2012 15:11
Edited by: user540519 on 19.12.2012 13:50

Hello
I ran the testcase here and reproduced the issue on 32 bit unmanaged beta v4.0.30319
note: same testcase works with Oracle.DataAccess (but not with managed).
This appears to match unpublished defect Bug 14666093 and is meant to be fixed in a later beta release.
Some things I noticed when testing
with the default fetchsize the breaking point is 67 iterations.
e.g.
private static int _intPara = 66;    // Works
private static int _intPara = 67; // Fails
If I increase the fetchsize then it breaks at different values as you noticed..
Hope this helps.
Kind Regards
John

1......100 from dual

Similar Messages

Maybe you are looking for