•Websheets
•Team Development
•Dynamic Actions
•Plug-Ins (Item & Region)
•Improved Charting including Gantts and Maps
•Declarative Tabular Forms
•REST Web Services
•Enhanced Interactive Reports
•Flexible Builder Authentication
•APEX Listener (currently in early adopters)
•Integrated jQuery and jQuery UI
•Improved Item Properties
•Additional Attributes
   Textareas: Resizable
   Combo Box (editable select list)
   Javascript Date pickers
   Autocomplete
•Integrated Oracle EBS Authentication
•APEX Advisor
•Editing Page Zero shows all application level components
•Improved UI Defaults with bi-directional synchronization
•Improved Themes for better out of the box user interfaces (xhtml)
•Support for native Excel Upload

There is still a lack of possibilities to create a table from scratch from a csv (with 1 line contains the column names).

The best method so far i found is:

1. Import csv into an access db (yes i know, but until you guys come up with a better idea).
Here you will have an access table with column names and datatypes (and your data)

2. Export this table by usng export => into an ODBC DB (which is your Oracle DB)

3. connect and you’ll have your Oracle table, with columns, datatypes and data

Quite easy and fast, if somebody has a better and faster idea (by not passing by an access db), let me know.

oracle-bi-ee-101332-url-parameters

promptness_please.html

In most case NOLOGGING generetaes no redo; true. But by adding a foreign key constraint to the table it disables direct mode inserts and resorts to conventional mode logging.

The direct load is also disabled for index organized table.

Managemnet and data collection activity can be suspended by imposing a blackout. See here some examples:

agentctl start blackout # Blackout the entrire agent
agentctl stop blackout # Resume monitoring and management
agentctl start blackout ORCL # Blackout DB ORCL
agentctl stop blackout ORCL # Resume monitoring DB ORCL
agentctl start blackout -s jobs -d 00:20 # Blackout jobs for 20 min

What somebody at Oracle Belgium told me once 10 years ago, I retested it on Oracle 10R2.
A Select ‘X’ is faster then a Select 1.
Stop dreaming because it’s only slightly faster.
see here the tests I did

SET time ON
SET timing ON
 
declare
 v_n number;
 v_c varchar2(1);
begin
FOR i IN 1..1000000 loop
 SELECT 'X' INTO v_c FROM dual;
end loop;
end;
/
 
declare
 v_n number;
 v_c varchar2(1);
begin
FOR i IN 1..1000000 loop
 SELECT 1 INTO v_n FROM dual;
end loop;
end;
/

And see here the results:
TRY    SELECT ‘X’       SELECT 1
___    __________     _________
1          27s00             27s03
2          27s00             27s02
3          26s09             27s02
4          27s00             27s03
5          27s00             27s04
6          53s08             55s00    (case 6 is with 2M itterations)

Had a query which took ages. After checking the session browser I noticed the query was doing a lot of “direct path write temp”. So I checked directly the sort_area_size. Big surprise it had a value of 64K!!!
After sending a mail to the DBA’s of my current project to ask to increase to 1GB, everything went smooth.

If you want to get rid of the content of USER_TAB_MODIFICATIONS or DBA_TAB_MODIFICATIONS then you have to execute dbms_stats.FLUSH_DATABASE_MONITORING_INFO

If that is taking a very long time then you might have hit the oracle bug 5709414