Oracle 10.2.0.4.0 Upgrade.

Hi All,
I have my databases running on Oracle 10.2.0.3.0 versions, can any one please let me know as to why or is it required to upgrade to 10.2.0.4.0.
Are there too many differences in both this releases.
Thanks in advance,
Regards,

[email protected] wrote:
can any one please let me know as to why or is it required to upgrade to 10.2.0.4.0. It is not required. If you currently don't have any issues then I would recommend to leave your system as it is. Never change a running system. But if you have particular issues and/or you want to request support from Oracle then it's a good idea to have the latest patch set installed, because it fixes a lot of bugs and very likely you will get one-off patches most easily for the latest patch set. If you're still on 10.2.0.3, you might get the answer that Oracle can't/won't provide a back-port to this release any longer and you need to install the latest patch-set first.
Are there too many differences in both this releases.I've attempted to provide a summary of the most crucial changes/new features of 10.2.0.4 a while ago:
http://oracle-randolf.blogspot.com/2008/04/overview-of-new-and-changed-features-in.html
Regards,
Randolf
Oracle related stuff blog:
http://oracle-randolf.blogspot.com/
SQLTools++ for Oracle (Open source Oracle GUI for Windows):
http://www.sqltools-plusplus.org:7676/
http://sourceforge.net/projects/sqlt-pp/

Similar Messages

Oracle 8.1.7.2.1 upgrade to Oracle 9

Oracle 8.1.7.2.1 upgrade to Oracle 9
I'm doing a manual migration to Oracle 9
When I do a shutdown immediate, and use the datafiles to do a clone, I find the database is not consistent when I do a Startup Migrate.
Is this a problem with this version of Oracle 8? Startup migrate will not let me do a open resetlogs...anyone got any suggestions thanks

Jeremy
From your question
- "Is this a problem with this version of Oracle 8? Startup migrate will not let me do a open resetlogs" -
what I guess you are doing is after cloning, you have tried starting Oracle using the same old Oracle 8.1.7 home. You need to make sure that that you change your oracle home to the new 9i home after you have cloned your files.
Also ensure that all required parameters are changes in the initialisation parameter (e.g COMPATIBLE) and those that relate to directories have changed to new. Also that all orace ENV (ORACLE_HOME, PATH, ORA_NLS33, LD_LIBRARY_PATH) point to the new 9i home.
Then try the STARTUP MIGRATE and run the upgrade script.

Oracle Follow-Up activities after EHP4 Upgrade

Hi!
I am currently executing the EHP4 Upgrade for SAP ERP 6.0 with Oracle.
After doing so I need to execute several oracle specific actions
Can some one give me technical documentation for the following actions:
1) Delete all empty Tablespaces
Question: How (SQL commando) can I identify which of my tablespaces are empty?
2) Tablespace "PSAPSR3700" should be released/deallocated
Question: what does it mean?
3) Load generation ???
Any helpful information will be very appreciated
Jürgen

> 1) Delete all empty Tablespaces
> Question: How (SQL commando) can I identify which of my tablespaces are empty?
You can check that using brtools.
> 2) Tablespace "PSAPSR3700" should be released/deallocated
> Question: what does it mean?
The sources are in PSAPSR3701 with EHP4 so that old tablespace is no more needed.
> 3) Load generation ???
Yes -run SGEN.
Markus

From wehre can I download Oracle Clusterware 11.1.0.7 (upgrade)

I need to upgrade my Clusterware installation to version 11.1.0.7, but I can't find a place from where to download it (Google doesn't help).
I've installed Oracle Clusterware 11g Release 1 (11.1.0.6.0) for Linux x86-64 from this page:
http://www.oracle.com/technetwork/database/enterprise-edition/downloads/111060-linx8664soft-099033.html
Please, if you know, tell me where can I find Clusterware 11.1.0.7 (for Redhat 5.4 64x).
Thank you.

Yes, you need a valid CSI (customer support identifier) to access the metalink support.
Regards
Asif Kabir
-- If you got the answer close the thread as well.

Oracle 10g OLAP to 11g OLAP upgrade ?

We currently are planning on a upgrade from 10g OLAP => 11g OLAP. We currently have 12 AWM's in 10g OLAP which we need to move over with associated DML programs and additional SQL reporting views.
Questions:
1. Is there any documentation available on necessary steps for 10g OLAP => 11g OLAP upgrade?
2. What would happen to existing AWM's prepared in 10g, would they be migrated to 11g transparently or have to be re-created?
3. Is there any specific documentation related to changes in way cube builds are done programatically?
4. Any changes in the way limit maps work in 11g?
Please advise.
Thanks,
Sudip

Migrating a 10g cube to 11g depends on whether you are talking about 11gR1 or 11gR2. 10g cubes continue to operate in the "10g way" even after a database upgrade. They will not become "11g cubes" until they are rebuilt after the database upgrade. In 11gR2, there is a supported way to migrate 10g cubes to 11g cubes, both with AWM and with PL/SQL. Sorry to say... this functionality doesn't exist in 11gR1: you'll have to rebuild your cubes from the ground up.
The SQL relational views built using the AWM plugin in 10g are no longer applicable in 11g. That's because OLAP cubes in 11g are registered in the oracle data dictionary (just like other Oracle objects), and the SQL relational views are managed in the database and a recognized part of the product. The SQL relational views are quite different in 11g, so you will likely have to rewrite queries against them.
LIMIT map syntax is the same, but performance is much better.
See if this blog entry helps:
http://www.rittmanmead.com/2009/10/09/olap-10gr2-and-dense-looping/
Edited by: Stewart Bryson on Feb 19, 2010 8:40 AM

Oracle 10g XE to 11g XE upgrade: gen_inst.sql does nothing.

I am trying to upgrade from 10gXE to 11gXE;
I logon to SQL Plus "conn sys as sysdba", and execute the gen_inst.sql procedure.
Quote: Connect to 10.2 XE database as SYS user and run gen_inst.sql. This will generate install.sql, gen_apps.sql and other .sql files. The files will be generated in the folder containing gen_inst.sql. End Quote.
However the procedure completes immediately, and does not generate any of the files described.
If I execute @(folder path)\fred.sql (where fred.sql does not exist, I get a message that the sql proc is not found, so I know that it finds the procedure when I try and execute it.
The following is a copy of the command window.
SQL*Plus: Release 10.2.0.1.0 - Production on Sat Jan 14 11:24:06 2012
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Enter user-name: sys as sysdba
Enter password:
Connected to:
Oracle Database 10g Express Edition Release 10.2.0.1.0 - Production
SQL> @c:\temp\gen_inst.sql
SQL> @c:\temp\fred.sql
SP2-0310: unable to open file "c:\temp\fred.sql"
SQL>None of the defined files , "install.sql, gen_apps.sql and other .sql files" are created.
If it try and run gen_inst.sql from within SQLDeveloper (2.1.1.64), after haveing connected as sys/sysdba, the procedure runs for several minutes, and creates some of the listed output files, before it finally crashes with :
Error report:
ORA-20000: ORU-10027: buffer overflow, limit of 1000000 bytes
ORA-06512: at "SYS.DBMS_OUTPUT", line 32
ORA-06512: at "SYS.DBMS_OUTPUT", line 97
ORA-06512: at "SYS.DBMS_OUTPUT", line 112
ORA-06512: at line 5
20000. 00000 - "%s"
*Cause: The stored procedure 'raise_application_error'
was called which causes this error to be generated.
*Action: Correct the problem as described in the error message or contact
the application administrator or DBA for more information.How do I get the script to run correctly in SQL Plus?
Edited by: Ian Mc on Jan 14, 2012 2:00 AM

Udo wrote:
I might just be a problem with spooling result files to the wrong location or calling them with the wrong path, because the script anticipates to have everything in the same local path...Interesting point, and definitely worth a try.
The quote part in op, from relevant doc section, seem to say differently. Looking through the script it would seem that files are actually spooled in working path (e.g. current dir when sqlplus is started) instead of folder where script exists.
"The files will be generated in the folder containing gen_inst.sql."
http://docs.oracle.com/cd/E17781_01/install.112/e18802/toc.htm#CIHCDCDH
I think file permissions should be considered still, if on Win 7 (or other OS with some level of least privilege policy implementation).
Edit:
A small test on windows platform showed that files are output in cwd from where sqlplus is started and also that saving files in c:\temp worked fine.
Edited by: orafad on Jan 15, 2012 12:40 AM

Oracle 10g RAC to 11g RAC Upgrade on Solaris

Hi,
We are planning to do a migration of a 4 Node Oracle 10g RAC on Solaris 9 to 11g with Solaris 10. We'd like know what would be the best path to take. We cannot afford any downtime!
Options: Are these feasible? Which option is best? Any documents links?
a) Do a rolling upgrade of Oracle from 10g to 11g. Then take down individual nodes and upgrade the Solaris OS from 9 to 10 and bring them up back into the cluster. Is there any known issues taking this path? Is a rolling upgrade like this possible?
b) Do an upgrade of the Solaris OS from 9 to 10 on each node and then bring them back up? Is this practical? Does Oracle allow different versions of OS running on different nodes?
c) Use Dataguard with 2 different RAC environments (2 nodes each). How would this work? Is it the only possible way? Any steps please?
Thanks

a) Do a rolling upgrade of Oracle from 10g to 11g. Then take down individual nodes and upgrade the Solaris OS from 9 to 10 and bring them up back into the cluster. Is there any known issues taking this path? Is a rolling upgrade like this possible?Hi,
first of all i would not change several components (OS, database version) at a time. My recommendation is to make small steps and start with the operating system first. Seconds recommendation is to test and test everything in your dev or test environment prior doing the upgrades in the productive environment. Trust me: You will face problems :-) So you better try it beforehand!
b) Do an upgrade of the Solaris OS from 9 to 10 on each node and then bring them back up? Is this practical? Does Oracle allow different versions of OS running on different nodes?As far i know you can run different operating system versions on different nodes if they are supported (Solaris 9 and 10 are).
Ronny Egner
My blog: http://ronnyegner.wordpress.com

Oracle Report Failing After Database Version Upgrade

Hi,
I have function code created in RDF file of oracle reports 10g as below-
create or replace function clobblobtest (TMP_MID varchar2) return varchar is
missed_text varchar2(1000) := '';
v_time_stamp date;
v_end_time date;
cursor c1(c_time_stamp in date,
          c_mid in varchar2) is
       select mn.TEXT from msgnotes mn
where mn.mid=c_mid
      and mn.create_date >= c_time_stamp;
cursor c2(c_time_stamp in date,
          c_mid in varchar2,
          c_end_time in date) is
       select mn.TEXT from msgnotes mn
where mn.mid=c_mid
      and mn.create_date >= c_time_stamp
      and mn.create_date <= c_end_time;
begin
IF ('1' = 'INWARD RETURN') then
    select bbi into missed_text from mtf1000 where mid=TMP_mid;
    ELSIF '1'='QPH Missed' then
Select max(nj.update_date) into v_time_stamp
from newjournal nj
where nj.mid=TMP_mid and nj.status = 'CTOVRQ';
for i in c1(v_time_stamp,TMP_mid) loop
missed_text:=missed_text||' '||i.text;
END Loop;
ELSIF ('Rejected'='Rejected' OR 'REJECTED'='REJECTED') then
IF ('TERMINATED' in ('TERMINATED')) THEN
Select min(nj.update_date),max(nj.ENDDATE) into v_time_stamp,v_end_time
from newjournal nj
where nj.mid=TMP_mid and nj.actionid2='16' and nj.actionid1 is null;
for i in c2(v_time_stamp,TMP_mid,v_end_time) loop
missed_text:=missed_text||' '||i.text;
END Loop;
ELSE
Select max(nj.update_date) into v_time_stamp
from newjournal nj
where nj.mid=TMP_mid and nj.status = 'UNSETLED';
for i in c1(v_time_stamp,TMP_mid) loop
missed_text:=missed_text||' '||i.text;
END Loop;
END IF;
ELSE missed_text := null;
END IF;
return(missed_text);
End;
Note:- Text column from table msgnotes is of datatype CLOB.
I have executed report having this function before oracle upgrade and it works fine.
But when i am trying to execute report after oracle version upgrade from 11.2.0.3.0 to 11.2.0.4.0 with same data report fails with runtime error Rep-1401: '' Fatal PL/SQL Error.
With New version we tried to typecast Text column from clob to varchar and executed report then it works... but report runs on old version without any explicit typecasting.
Please note when i run and call this funtion from sql developer with upgraded database version it works fine.
Please advise.

This error can possibly depend on the fact that the PL/SQL engines are in different versions. When using SQL Developer you use the PL/SQL engine in the database. When using Oracle Reports you use the PL/SQL engine in the Oracle Reports Server (Web Logic Server even maybe). These versions - in the Forms and Reports realm - are different (read: a lower version) than the PL/SQL engine in the database.
If I check what version I have in Oracle Forms, as an example, I get 11.1.0.7.0 (Help in the Oracle Forms Builder menu) but in the database I have (select * from product_component_version;) 11.2.0.3.0.
That COULD be your issue...how to fix it I am not sure...

Oracle AQ tables - 9i to 10g upgrade

We have Oracle AQ implemented in one of our application. Our database on the test environment recently went through an upgrade from 9i to 10g. Now when we deploy the application we are getting the following error:
Caused by: oracle.oc4j.admin.internal.DeployerException: Error initializing ejb-module; Exception Error in application mdo: Error loading package at file:/oc4j/oc4j10.1.2.0/j2ee/soc4j/applications/mdo/mdo.jar, Error deploying file:/oc4j/oc4j10.1.2.0/j2ee/soc4j/applications/mdo/mdo.jar homes: JMS Error: ORA-01031: insufficient privileges
ORA-06512: at "SYS.DBMS_AQADM_SYS", line 5989
ORA-06512: at line 1
ORA-06512: at "SYS.DBMS_AQJMS", line 174
ORA-06512: at line 1
The user does have the execute privileges. A quick look at the queue table revealed that the table has changed significantly after the 9i to 10g database upgrade and that could be the reason we are seeing this error. Quick differences we are seeing are all columns with DATE datatype on 9i are now defined as TIMESTAMP(6) under 10g, and there is an additional column call user_prop of type SYS.ANYDATA.
Can somebody please confirm this ? I thought Oracle AQ will be backwards compatible and will be an abstract interface to the application but looks like its not. How do I get this issue resolved ?

I suggest you revoke all the privileges and grant once again.
Privileges:-
GRANT EXECUTE ON DBMS_AQ TO User;
GRANT EXECUTE ON DBMS_AQADM TO User;
GRANT EXECUTE ON DBMS_AQJMS TO User;
GRANT EXECUTE ON DBMS_AQIN TO User;
Regards,
Sanjeev

AIX 64 Bit - Oracle 11.2.0.3 Installation/Upgrade - 10.2/11.1

AIX 64 Bit
Oracle 11gR2 - 11.2.0.3
As part of the 11gR2 install, you are asked to select DBA and DBOPER owners. How is this determined/verified selected properly?
Also, migrating DB's: Ran the dbupgdiag.sql and utlu112i_3.sql
Should we enable 11.2 listener first?
or
Should we migrate all DB's, then do listener or can we try one DB with new listener then move the rest and then change to the 11.2.0.3 LSNR?
Is there a best practice here?

Hi;
As part of the 11gR2 install, you are asked to select DBA and DBOPER owners. How is this determined/verified selected properly?Its already mention at installation guide. What is purpose of those groups, what you should choose etc.
Also, migrating DB's: Ran the dbupgdiag.sql and utlu112i_3.sql
Should we enable 11.2 listener first?
or
Should we migrate all DB's, then do listener or can we try one DB with new listener then move the rest and then change to the 11.2.0.3 LSNR?You should go wiht first db upgrade. Than you can go wiht listener. As suggestion here by orafad that issue is related wiht db upgrade soo make post on related forum site
Regard
Helios

Installing Oracle 10.2.0.1 and upgrade to 10.2.0.3 on WIndows 2000 using VM

What's up with trying to install Oracle 10.2.0.1 database standard edition on a Windows 2000 OS which utilizes VMWARE? This has been the install from hell !!
I just did the whole process on my Windows XP laptop in about 1 hour and the VMWARE Windows 2000 install/upgrade is on its third day and we have not even migrated the data yet.
Has anyone done such an upgrade on the same Windows 2000 OS using VMWARE? The server configuration sucks as well with only one CPU, small processor and and little over 1 gb of memory but this is what VMWARE is allocating I assume. The customer also wants all his database files on the SAN. Should I tell the management they need more resources or is this upgrade possible????

Not if it isn't required to create a secure and stable environment.
What I am referring to is an operating system that doesn't require an anti-virus client.
And operating system that doesn't force you to support a massive GUI that no one will ever look at after installation is completed.
An operating system that doesn't have a vendor that admits to violating compliance and altering system files without the owner's permission.

Oracle 11.1.0.6 RDBMS Upgrade to 11.2 What Patch Number for Linux?

I have Oracle RDBMS 11.1.0.6 on 64bit Linux two node RAC and trying to find out what is the patch number to upgrade the database to R2. Thanks

yakub21 wrote:
I have Oracle RDBMS 11.1.0.6 on 64bit Linux two node RAC and trying to find out what is the patch number to upgrade the database to R2. ThanksYakub,
Patches are only used within a release. 11.2 is '11g Release 2' - a new release. This is no longer a patch, it is an upgrade.
You perform the upgrade as per the Upgrade manual in the doc set. PLEASE read that manual before you attempt te upgrade. There are a few steps and concepts that are significantly different from a patch.
The starting point for an upgrade is the freely downloaded s/w. The software is a completely fresh install, but the database gets upgraded.
By the way, I would not recommend jumping on 11.2 for an existing production environment until a careful evaluation has been made - upgrades always introduce significant changes and they could adversely affect an existing application is not properly tested and understood. Budget a few weeks (at least) to understand the differences.

Oracle packages - nested table cursor - 11g upgrade testing

I have a package call Crisis_Contacts. This package has a function(generate_file) in it that has a create cusor statement which calls on another function(split) in another package called NCF_UTL. In testing our code for a 11g upgrade the cursor statement calling on the split function is throwing a Error: ORA-00600: internal error code, arguments: [4814], [5], [0], [0], [], [], [], [], [], [], [], [] when compiling the package, which it was not doing prior to upgrade. Can you suggest a possible workaround or mod to this code to stop the error from occuring.
Basically, this code is reading html files and spliting them into section and loading them into the cursor. Then the cursor is read to write them to a table. Then it reads the table and transmits the lines of html to another location.
CREATE OR REPLACE package body ACADMGR.crisis_contacts as
function generate_file( file_type in varchar2,
person_pidm in integer := null,
person_id in varchar2 := null,
dry_run in varchar2 := 'N',
pidm_filter in varchar2 := null,
include_placeholders in varchar2 := 'N',
comments in varchar2 := null ) return integer is
cursor typeDef is
select * from crisis_contact_file_type ccft
where ccft.file_type = generate_file.file_type;
file_type_def typeDef%rowtype;
cursor fieldDef is
select * from crisis_contact_file_field ccff
where ccff.file_type = generate_file.file_type
order by seq asc;
type r_typ is ref cursor;
r r_typ;
datasel varchar2(32767) := '';
r_pidm integer;
pers_pidm integer := person_pidm;
prev_pidm integer;
r_tag varchar2(255);
r_val varchar2(4000);
type field_list is table of varchar2(256);
type field_index is table of integer index by varchar2(256);
type val_list is table of varchar2(4000);
type req_tbl is table of boolean index by varchar2(256);
req_fields field_list := field_list();
all_req_seen boolean;
fields field_list := field_list();
vals val_list := val_list();
field_idx field_index;
out_id integer;
out_line integer := 0;
buf varchar2(4000);
saved_out_line integer := 0;
xml_section_lf_delims ncf_utl.split_tbl := ncf_utl.split_tbl( ncf_utl.crlf, chr(13), chr(10) );
cursor splitcur( txt in varchar2 ) is
select column_value line from table(ncf_utl.split( txt, xml_section_lf_delims ));
cursor get_output is
select text from crisis_contact_file_line
where output_id = out_id
order by line_no asc;
log_time date := sysdate;
log_seq integer := 1;
procedure outp( txt in varchar2 ) is
begin
--dbms_output.put_line( txt );
insert into crisis_contact_file_line
( output_id, line_no, text )
values
( out_id, out_line, txt );
out_line := out_line + 1;
end;
procedure split_outp( txt in varchar2 ) is
begin
if txt is not null then
for r in splitCur( txt ) loop
outp( r.line );
end loop;
end if;
end;
begin
open typeDef;
fetch typeDef into file_type_def;
close typeDef;
if pers_pidm is null and person_id is not null then
select distinct(spriden_pidm) into pers_pidm from spriden
where spriden_id = person_id;
end if;
insert into crisis_contact_file_out ( file_type, comments ) values ( file_type, comments )
returning output_id into out_id;
datasel := gen_outfile_sql( file_type, pers_pidm, pidm_filter, include_placeholders );
if dry_run = 'Y' then
ncf_utl.output_text( datasel );
else
split_outp( file_type_def.preamble );
buf := '';
for f in fieldDef loop
fields.extend;
fields(fields.last) := f.alias;
field_idx(f.output_field) := fields.last;
vals.extend();
if f.mandatory = 'Y' then
req_fields.extend;
req_fields(req_fields.last) := f.output_field;
end if;
if file_type_def.file_mode = 'DELIM' and file_type_def.print_heading = 'Y' then
buf := buf || case when fieldDef%rowcount > 1 then nvl(file_type_def.delimiter,',') else '' end ||
nvl(file_type_def.quote,'') || nvl(f.alias,f.output_field) || nvl(file_type_def.quote,'');
end if;
end loop;
if length(buf) > 0 then
outp(buf);
end if;
--outp( '<?xml version="1.0" encoding="UTF-8"?>' );
--outp( '<recipients>' );
open r for datasel;
loop
fetch r into r_pidm, r_tag, r_val;
exit when r%notfound;
if prev_pidm is null or prev_pidm != r_pidm then
if prev_pidm is not null then
split_outp( file_type_def.record_end );
all_req_seen := true;
--check and reset 'seen' flags for required fields
for i in 1..req_fields.count loop
if vals(field_idx(req_fields(i))) is null then
all_req_seen := false;
end if;
end loop;
--if not all_req_seen then
-- rollback to USER_START;
-- out_line := saved_out_line;
--end if;
if all_req_seen then
buf := '';
for i in 1..fields.count loop
if file_type_def.file_mode = 'XML' then
outp( ' <' || fields(i) || '>' || vals(i) || '</' || fields(i) || '>' );
elsif file_type_def.file_mode = 'DELIM' then
buf := buf || case when i > 1 then nvl(file_type_def.delimiter,',') else '' end ||
nvl(file_type_def.quote,'') || vals(i) || nvl(file_type_def.quote,'');
end if;
end loop;
if file_type_def.file_mode = 'DELIM' then
outp(buf);
end if;
end if;
end if;
--savepoint USER_START;
--saved_out_line := out_line;
if r_tag != 'endofdata' then
split_outp( file_type_def.record_start );
end if;
for i in 1..fields.count loop
vals(i) := null;
end loop;
end if;
prev_pidm := r_pidm;
if field_idx.exists(r_tag) then
if r_val is not null then
vals(field_idx(r_tag)) := r_val;
end if;
end if;
end loop;
if pers_pidm is not null then
split_outp( file_type_def.record_end );
end if;
split_outp( file_type_def.postamble );
--for o in get_output loop
-- dbms_output.put_line( o.text );
--end loop;
end if;
return out_id;
end generate_file;
procedure gen_and_transmit( file_type in varchar2,
username in varchar2, pass in varchar2,
pidm_filter in varchar2 := null,
include_placeholders in varchar2 := 'N',
comments in varchar2 := null,
url_base in varchar2 := null, url_path in varchar2 := null,
ssl_wallet_path in varchar2 := null, ssl_wallet_pass in varchar2 := null ) is
out_id integer;
begin
out_id := generate_file( file_type, pidm_filter => pidm_filter,
include_placeholders => include_placeholders,
comments => comments );
--transmit_file( out_id, username, pass,
transmit_file( out_id, 'mhaywood', '------------',
url_base => url_base, url_path => url_path,
ssl_wallet_path => ssl_wallet_path, ssl_wallet_pass => ssl_wallet_pass );
end gen_and_transmit;
begin
UTL_HTTP.set_persistent_conn_support( true, 10 );
end;
create or replace
PACKAGE BODY NCF_UTL IS
PROCEDURE OUTPUT_TEXT( BUFFER IN VARCHAR2, LINE_LEN IN INTEGER := 80 ) IS
TYPE CHRLIST IS TABLE OF VARCHAR2(1);
cr VARCHAR2(1) := CHR(13);
     lf VARCHAR2(1) := CHR(10);
     tab VARCHAR2(1) := CHR(9);
     whitespace CHRLIST := CHRLIST( ' ', tab );
     pos INTEGER := 1;
     len INTEGER;
     brk INTEGER;
     curbrk INTEGER;
     lnbrk INTEGER;
     whtbrk INTEGER;
     i INTEGER;
BEGIN
len := LENGTH(BUFFER);
     WHILE pos <= len
     LOOP
     WHILE SUBSTR( BUFFER, pos, 1 ) IN ( cr, lf )
     LOOP
     pos := pos + 1;
     END LOOP;
     --find the next newline type char, or pos+250 if one isn't found
lnbrk := LEAST( NVL(NULLIF(INSTR( BUFFER, cr, pos ),0),pos+250),
                         NVL(NULLIF(INSTR( BUFFER, lf, pos ),0),pos+250) );
     --after the loop:
     -- whtbrk should be the last whitespace char before LINE_LEN + pos
     -- (or == pos, if none found)
     -- curbrk should be == 0 or the 1st whitespace after LINE_LEN + pos
     whtbrk := pos;
     curbrk := pos;
     FOR i IN 1..whitespace.COUNT
     LOOP
     LOOP
     curbrk := INSTR( BUFFER, whitespace(i), whtbrk + 1 );
          IF curbrk <= pos + LINE_LEN AND curbrk >= whtbrk THEN
          whtbrk := curbrk;
          END IF;
          EXIT WHEN curbrk > pos + LINE_LEN OR curbrk = 0;
          END LOOP;
          EXIT WHEN curbrk > pos + LINE_LEN;
     END LOOP;
     brk := LEAST( len + 1, lnbrk, NVL( NULLIF( whtbrk, pos ), pos + 250 ),
                    NVL( NULLIF( curbrk, 0 ), pos + 250 ), pos + 250 );
     dbms_output.put_line( SUBSTR( BUFFER, pos, brk - pos ) );
     pos := brk;
     END LOOP;
END;
function split( str in varchar2, d0 in varchar2, d1 in varchar2 := null,
d2 in varchar2 := null, d3 in varchar2 := null,
d4 in varchar2 := null, d5 in varchar2 := null,
d6 in varchar2 := null, d7 in varchar2 := null,
d8 in varchar2 := null, d9 in varchar2 := null )
return split_tbl pipelined is
pos integer := 1;
curidx integer;
idx integer;
delims split_tbl;
idx_delim varchar2(32767);
len integer := length( str );
begin
if d0 is null then
--split on whitespace
delims := split_tbl( ' ', chr(9), chr(10), chr(13) );
else
delims := split_tbl( d0, d1, d2, d3, d4, d5, d6, d7, d8, d9 );
end if;
loop
idx := -1;
--we want the lowest value of IDX > 0
for i in 1..delims.count loop
curidx := instr( str, delims(i), pos, 1 );
if curidx > 0 then
if idx = -1 or curidx < idx then
idx := curidx;
idx_delim := delims(i);
end if;
end if;
end loop;
if idx > 0 then
--if splitting on whitespace, treat any amount of it as
--a single delimiter
if d0 is not null or idx > pos then
pipe row( substr( str, pos, idx - pos ) );
end if;
else
-- also make sure to ignore trailing whitespace
if d0 is not null or pos <= len then
pipe row( substr( str, pos ) );
end if;
exit;
end if;
pos := idx + length( idx_delim );
end loop;
return;
end split;
function split( str in varchar2, delimiters in split_tbl := null )
return split_tbl pipelined is
i integer;
cursor spcur( d0 in varchar2 := null, d1 in varchar2 := null,
d2 in varchar2 := null, d3 in varchar2 := null,
d4 in varchar2 := null, d5 in varchar2 := null,
d6 in varchar2 := null, d7 in varchar2 := null,
d8 in varchar2 := null, d9 in varchar2 := null ) is
select column_value v from table (split(str,d0,d1,d2,d3,d4,d5,d6,d7,d8,d9));
function x( i in integer ) return varchar2 is
begin
if delimiters.exists(i) then
return delimiters(i);
else
return null;
end if;
end;
begin
for r in spcur( x(0),x(1),x(2),x(3),x(4),x(5),x(6),x(7),x(8),x(9) ) loop
pipe row( r.v );
end loop;
end split;
END;
Thanks

Curious behavior, but irrelevant, IMHO, since one never switches back and forth between two releases in the real world. All that STARTUP UPGRADE does is that it enables upgrade scripts to be run - it does not prevent statements like CREATE USER from being executed since the upgrade may create new needed accounts
http://docs.oracle.com/cd/E11882_01/server.112/e16604/ch_twelve045.htm#SQPUG128
HTH
Srini

Oracle 11g R1(11.1) Database Upgrade Guide.

Where can I find the SAP documentation for:
Database Upgrade guide > Upgrade to Oracle Database 11g Release 1(11.1):UNIX.
I am able to find Upgrade to Oracle Database 11g Release 2(11.2). We are planning to upgrade to 11g Release 1(11.1) and I am not able to find any SAP documentation for 11g Release 1(11.1):
Does SAP support 11g Release 1(11.1)?
We are using Business Object XIR3 and want to upgrade from Oracle 9i to Oracle 11g R1.
Thanks
Edited by: Sherry Barkodar on Jan 7, 2011 1:52 AM

We are using Oracle 9i for our Business Object repository for CMS / Audit repository purpose and we want to upgrade to Oracle 11g. What is the impact of this upgrade (from Oracle 9i to Oracle 11g) on Business Object repository?
Can you please provide some information for upgrade to Oracle 11g R2 and impact on business Object Repository?
Bellow is information about our platform:
-          The BOE XI 3.1 installation is on Sun Solaris.
-          Upgrading from Oracle 9i to Oracle 11g to be used for CMS / Audit repository purpose.
-          We want to use Oracle 11g for CMS repository and reporting purposes.
Thanks,
Sherry

Oracle 10.2.0.4 patch upgrade

Hi,
We have recently upgraded oralce 10.2.0.2 to 10.2.0.4 we downloaed the Oracle patches as per snote # 871735 for SUN aparc 64 bit .
We have used OUI and done the upgrade as per the instaructions in the readme file of the patch.
I have observed in Snote #1137346 that after upgrading to 10.2.0.4 apply the pacthes as per snotes #1137346.
Environment:
R/3 4.6c and BW 7.0
OS : Sun sparc 5.9 64 bit
DB : Oracle 10.2.0.4
Please suggest does applying the oracle patches as per snote # 871735 is enough or we have to go as per#1137346 also....
Thanks,
Subhash.G

> We have recently upgraded oralce 10.2.0.2 to 10.2.0.4 we downloaed the Oracle patches as per snote # 871735 for SUN aparc 64 bit .
> We have used OUI and done the upgrade as per the instaructions in the readme file of the patch.
Ok.
> I have observed in Snote #1137346 that after upgrading to 10.2.0.4 apply the pacthes as per snotes #1137346.
> Please suggest does applying the oracle patches as per snote # 871735 is enough or we have to go as per#1137346 also....
Yes - you also have to apply all the interim patches too. You updated your database only to 10.2.0.4; this version has also bugs which are fixed byy applying all the patches of note 1137346.
Markus

Oracle 10.2.0.4.0 Upgrade.

Similar Messages

Maybe you are looking for