Oracle 11g PL/SQL notes

Agenda

Oracle PL/SQL usage and 11g new features for Developers for DBAs

• Row by row processing
• Nested row by row processing
• Lookup queries
• Result Cache for lookup queries
• Excessive access to DUAL
• Populating Master-Detail Rows
• Unnecessary Function Execution
• Excessive Parsing
• Excessive Commits
• SIMPLE_INTEGER Datatype
• Continue Statement
• Sequences in PL/SQL expressions
• Dynamic SQL Enhancements
• Native Compilation
• Automatic Subprogram Inlining
• READ ONLY Tables
• INVISIBLE indexes
• SQL_monitoring
• History Tables & Enterprise Manager
• References

Y. Anıl Akduygu October - 2012

Row-by-Row Processing
DO NOT USE USE
DECLARE --
CURSOR c1 IS -- Insert in to target table
SELECT prod_id, cust_id, time_id, amount_sold --
FROM salesWHERE amount_sold > 100;
c1_rec c1%rowtype;
INSERT INTO top_sales_customers (prod_id,
l_cust_first_name customers.cust_first_name%TYPE;
cust_id,
l_cust_lasT_name customers.cust_last_name%TYPE;
BEGIN time_id,
FOR c1_rec IN c1 cust_first_name,
LOOP cust_last_name,
-- Query customer details amount_sold)
SELECT cust_first_name, cust_last_name SELECT s.prod_id,
INTO l_cust_first_name, l_cust_last_name s.cust_id,
FROM customers s.time_id,
WHERE cust_id=c1_rec.cust_id; c.cust_first_name,
-- c.cust_last_name,
-- Insert in to target table
s.amount_sold
--
FROM sales s, customers c
INSERT INTO top_sales_customers (
prod_id, cust_id, time_id, cust_first_name, cust_last_name,amount_sold WHERE s.cust_id = c.cust_id AND s.amount_sold > 100;
)
VALUES
(
c1_rec.prod_id,
c1_rec.cust_id,
c1_rec.time_id,
l_cust_first_name,
l_cust_last_name,
c1_rec.amount_sold
);
END LOOP;
COMMIT;
END;

SQL statements are called from PL/SQLin a loop, so the execution will switch back and forth between the PL/SQL engine and the SQL engine.
This switch between two environments is known as a context switch. Context switches increase elapsed time of your programs

Nested Row-by-Row Processing
DO NOT USE USE

DECLARE MERGE INTO fact1 USING
CURSOR c1 AS (SELECT DISTINCT c3.n1,c3.n2
SELECT n1 FROM t1; FROM t1, t2, t3
CURSOR c2 (p_n1) AS WHERE t1.n1 = t2.n1
SELECT n1, n2 FROM t2 WHERE n1=p_n1; AND t2.n1 = t3.n1
CURSOR c3 (p_n1, p_n2) AS AND t2.n2 = t3.n2) t
SELECT text FROM t3 WHERE n1=p_n1 AND n2=p_n2; ON (fact1.n1=t.n1 AND fact1.n2=t.n2)
BEGIN WHEN matched THEN
FOR c1_rec IN c1 UPDATE SET .. WHEN NOT matched THEN
LOOP INSERT .. ;
FOR c2_rec IN c2 (c1_rec.n1) COMMIT;
LOOP ;
FOR c3_rec IN c3(c2_rec.n1, c2_rec.n2)
LOOP
-- execute some sql here;
UPDATE … SET ..where n1=c3_rec.n1 AND n2=c3_rec.n2;
EXCEPTION
WHEN no_data_found THEN
INSERT into… END;
END LOOP;
END LOOP;
END LOOP;
COMMIT;
END;
/

Do not write code with deeply nested cursors in PL/SQL language. Review it to see if you can write such code in SQL instead.

Lookup Queries
DO NOT USE USE
DECLARE
DECLARE CURSOR c1
CURSOR c1 IS
IS SELECT prod_id,cust_id,time_id,
SELECT prod_id, amount_sold
cust_id, FROM sales
time_id, WHERE amount_sold > 100;
amount_sold
FROM sales l_country_names country_names_type;
WHERE amount_sold > 100; l_Country_id countries.country_id%TYPE;
l_country_name countries.country_name%TYPE;
l_cust_first_name customers.cust_first_name%TYPE; l_cust_first_name customers.cust_first_name%TYPE;
l_cust_last_name customers.cust_last_name%TYPE; l_cust_lasT_name customers.cust_last_name%TYPE;
l_Country_id countries.country_id%TYPE;
l_country_name countries.country_name%TYPE; TYPE country_names_type
BEGIN IS
FOR c1_rec IN c1 TABLE OF VARCHAR2 (40)
LOOP INDEX BY PLS_INTEGER;
-- Query customer details
SELECT cust_first_name, cust_last_name, country_id l_country_names country_names_type;
INTO l_cust_first_name, l_cust_last_name, l_country_id BEGIN
FROM customers FOR c1_rec IN c1
WHERE cust_id = c1_rec.cust_id; LOOP
-- Query to get country_name SELECT cust_first_name, cust_last_name, country_id
SELECT country_name INTO l_cust_first_name, l_cust_last_name, l_country_id
INTO l_country_name FROM customers
FROM countries WHERE cust_id = c1_rec.cust_id;
WHERE country_id = l_country_id;
-- Check array first before executing a SQL statement
-- IF (l_country_names.EXISTS (l_country_id))
-- Insert in to target table THEN
-- l_country_name := l_country_names (l_country_id);
INSERT INTO top_sales_customers (prod_id, ELSE
cust_id, SELECT country_name
time_id, INTO l_country_name
cust_first_name, FROM countries
cust_last_name, WHERE country_id = l_country_id;
amount_sold, -- Store in the array for further reuse
country_name) l_country_names (l_country_id) := l_country_name;
VALUES (c1_rec.prod_id, END IF;
c1_rec.cust_id, -- Insert in to target table
c1_rec.time_id, INSERT INTO top_sales_customers
l_cust_first_name, (prod_id,cust_id,time_id,cust_first_name,
l_cust_last_name, cust_last_name,amount_sold,country_name)
c1_rec.amount_sold, VALUES
l_country_name); (c1_rec.prod_id,c1_rec.cust_id,c1_rec.time_id,l_cust_first_name,
END LOOP; l_cust_last_name,c1_rec.amount_sold,l_country_name);
END LOOP;
COMMIT;
END; COMMIT;
• / END;
/

You can define an associative array to cache the results of the lookup query and reuse the array in later executions, thus effectively reducing the
executions of the lookup query.

Result Cache for Lookup Queries
DO NOT USE USE
DECLARE
DECLARE CURSOR c1
CURSOR c1 IS
IS SELECT prod_id,
SELECT prod_id, cust_id,
cust_id, time_id,
time_id, amount_sold
amount_sold FROM sales
FROM sales WHERE amount_sold > 100;
WHERE amount_sold > 100;
l_cust_first_name customers.cust_first_name%TYPE;
l_cust_first_name customers.cust_first_name%TYPE; l_cust_last_name customers.cust_last_name%TYPE;
l_cust_last_name customers.cust_last_name%TYPE; l_Country_id countries.country_id%TYPE;
l_Country_id countries.country_id%TYPE; l_country_name countries.country_name%TYPE;
l_country_name countries.country_name%TYPE; BEGIN
BEGIN FOR c1_rec IN c1
FOR c1_rec IN c1 LOOP
LOOP -- Query customer details
-- Query customer details SELECT cust_first_name, cust_last_name, country_id
SELECT cust_first_name, cust_last_name, country_id INTO l_cust_first_name, l_cust_last_name, l_country_id
INTO l_cust_first_name, l_cust_last_name, l_country_id FROM customers
FROM customers WHERE cust_id = c1_rec.cust_id;
WHERE cust_id = c1_rec.cust_id;
-- Query to get country_name
-- Query to get country_name
SELECT country_name
SELECT /*+ RESULT_CACHE */ country_name
INTO l_country_name
INTO l_country_name
FROM countries
FROM countries
--
--
--
--
cust_id,
cust_id,
time_id,
time_id,
cust_first_name,
cust_first_name,
cust_last_name,
cust_last_name,
amount_sold,
amount_sold,
country_name)
country_name)
VALUES (c1_rec.prod_id,
VALUES (c1_rec.prod_id,
c1_rec.cust_id,
c1_rec.cust_id,
c1_rec.time_id,
c1_rec.time_id,
l_cust_first_name,
l_cust_first_name,
l_cust_last_name,
l_cust_last_name,
c1_rec.amount_sold,
c1_rec.amount_sold,
l_country_name);
l_country_name);
END LOOP;
END LOOP;
COMMIT;
COMMIT;
END;
END;
/
/

The result cache is new to Oracle 11g and provides enhanced query performance for SQL and PL/SQL applications by caching the results of SQL queries
into memory . A result cache shareable and is stored in SGA memory.

Excessive Access to DUAL
DO NOT USE USE

DECLARE DECLARE
l_epoch INTEGER; l_epoch INTEGER;
BEGIN BEGIN
SELECT ( (SYSDATE l_epoch :=
- TO_DATE ('01-JAN-1970 00:00:00', 'DD-MON-YYYY HH24:MI:SS')) (SYSDATE - TO_DATE ('01-JAN-1970 00:00:00', 'DD-MON-YYYY HH24:MI:SS'))
* 24 * 24
* 60 * 60
* 60) * 60;
INTO l_epoch
FROM DUAL; DBMS_OUTPUT.put_line (l_epoch);
END;
DBMS_OUTPUT.put_line (l_epoch); /
END;
/

You should avoid overusing DUAL table access. Accessing DUAL from PL/SQL causes context switching, which hurts performance.

Excessive Access to DUAL
DO NOT USE USE

DECLARE
l_cust_id NUMBER; INSERT INTO customers_hist
BEGIN SELECT cust_hist_id_seq.NEXTVAL, cust_first_name, cust_last_name
FOR c1 IN (SELECT cust_first_name, cust_last_name FROM customers
FROM customers WHERE cust_marital_status != 'married';
WHERE cust_marital_status != 'married')
LOOP
SELECT cust_hist_id_seq.NEXTVAL INTO l_cust_id FROM DUAL;

INSERT INTO customers_hist (cust_hist_id, first_name, last_name)
VALUES (l_cust_id, c1.cust_first_name, c1.cust_last_name);
END LOOP;
END;
/

You should avoid overusing DUAL table access. Accessing DUAL from PL/SQL causes context switching, which hurts performance.

Populating Master-Detail Rows
DO NOT USE USE

DECLARE
l_cust_id NUMBER; INSERT INTO customers (cust_id, ...)
BEGIN VALUES (cust_id_seq.nextval,...)
FOR c1 IN (SELECT cust_first_name, cust_last_name RETURNING cust_id into l_cust_id;
FROM customers
WHERE cust_marital_status != 'married') INSERT INTO customer_transactions (cust_id, ...)
LOOP VALUES (l_cust_id,...)
SELECT cust_hist_id_seq.NEXTVAL INTO l_cust_id FROM DUAL; ...

INSERT INTO customers_hist (cust_hist_id, first_name, last_name)
VALUES (l_cust_id, c1.cust_first_name, c1.cust_last_name);
END LOOP;
END;
/

You can retrieve the key value from a newly-inserted master row by using the DML RETURNING clause. Then you can use that key value while inserting in
to the detail table.

Unnecessary Function Execution
DO NOT USE USE
CREATE TABLE log_table (message_seq NUMBER, MESSAGE VARCHAR2 (512));
DECLARE
CREATE SEQUENCE message_id_seq; l_debug BOOLEAN := FALSE;
r1 INTEGER;
DECLARE
l_debug BOOLEAN := FALSE; FUNCTION log_entry (v_message IN VARCHAR2, v_debug IN BOOLEAN)
r1 INTEGER; RETURN NUMBER
IS
FUNCTION log_entry (v_message IN VARCHAR2, v_debug IN BOOLEAN) BEGIN
RETURN NUMBER IF (v_debug)
IS THEN
BEGIN INSERT INTO log_table (message_seq, MESSAGE)
IF (v_debug) VALUES (message_id_seq.NEXTVAL, v_message);
THEN END IF;
INSERT INTO log_table (message_seq, MESSAGE)
VALUES (message_id_seq.NEXTVAL, v_message); RETURN 0;
END IF; END;
BEGIN
RETURN 0; FOR c1 IN (SELECT s.prod_id,
END; s.cust_id,
BEGIN s.time_id,
FOR c1 IN (SELECT s.prod_id, c.cust_first_name,
s.cust_id, c.cust_last_name,
s.time_id, s.amount_sold
c.cust_first_name, FROM sales s, customers c
c.cust_last_name, WHERE s.cust_id = c.cust_id AND s.amount_sold > 100)
s.amount_sold LOOP
FROM sales s, customers c $IF $$debug_on
WHERE s.cust_id = c.cust_id AND s.amount_sold > 100) $THEN
LOOP IF c1.cust_first_name IS NOT NULL
IF c1.cust_first_name IS NOT NULL THEN
THEN r1 := log_entry ('first_name is not null ', l_debug);
r1 := log_entry ('first_name is not null ', l_debug); END IF;
END IF;
IF c1.cust_last_name IS NOT NULL
IF c1.cust_last_name IS NOT NULL THEN
THEN r1 := log_entry ('Last_name is not null ', l_debug);
r1 := log_entry ('Last_name is not null ', l_debug); END IF;
END IF; $END
END LOOP; NULL;
END; END LOOP;
/ END;
/

Executing a function call usually means that a different part of the instruction set must be loaded into the CPU. By avoiding unnecessary function
execution, you avoid unneeded flushing and refilling of the instruction pipeline, thus minimizing demands upon your CPU.

DO NOT USE USE
CREATE OR REPLACE FUNCTION calculate_epoch (d IN DATE) CREATE INDEX compute_epoch_fbi ON sales
RETURN NUMBER (calculate_epoch(time_id))
DETERMINISTIC PARALLEL (DEGREE 4);
IS
l_epoch NUMBER; SELECT
BEGIN MAX (calculate_epoch (s.time_id)) epoch
l_epoch := FROM sales s
(d - TO_DATE ('01-JAN-1970 00:00:00', 'DD-MON-YYYY HH24:MI:SS')) WHERE s.amount_sold > 100
* 24 AND calculate_epoch (s.time_id) BETWEEN 1000000000 AND 1100000000;
* 60
* 60;
RETURN l_epoch;
END calculate_epoch;
/

SELECT
MAX (calculate_epoch (s.time_id)) epoch
FROM sales s
WHERE s.amount_sold > 100
AND calculate_epoch (s.time_id) BETWEEN 1000000000 AND 1100000000;

Executing a function call usually means that a different part of the instruction set must be loaded into the CPU. By avoiding unnecessary function
execution, you avoid unneeded flushing and refilling of the instruction pipeline, thus minimizing demands upon your CPU.

DO NOT USE USE
CREATE OR REPLACE FUNCTION calculate_epoch (d IN DATE) • CREATE OR REPLACE FUNCTION calculate_epoch (d IN date)
RETURN NUMBER • RETURN NUMBER DETERMINISTIC RESULT_CACHE IS
DETERMINISTIC • l_epoch number;
IS • BEGIN
l_epoch NUMBER; • l_epoch := (d - TO_DATE('01-JAN-1970 00:00:00', 'DD-MON-YYYY HH24:MI:SS'))
BEGIN • * 24 *60 *60 ;
l_epoch := • RETURN l_epoch;
(d - TO_DATE ('01-JAN-1970 00:00:00', 'DD-MON-YYYY HH24:MI:SS')) • END calculate_epoch;
* 24 • /
* 60
* 60; • SELECT
RETURN l_epoch; • MAX (calculate_epoch (s.time_id)) epoch
END calculate_epoch; • FROM sales s
/ • WHERE s.amount_sold > 100
• AND calculate_epoch (s.time_id) BETWEEN 1000000000 AND 1100000000;
SELECT
MAX (calculate_epoch (s.time_id)) epoch
FROM sales s
WHERE s.amount_sold > 100
AND calculate_epoch (s.time_id) BETWEEN 1000000000 AND 1100000000;

The result cache is new to Oracle 11g and provides enhanced query performance for SQL and PL/SQL applications by caching the results of SQL queries
into memory . A result cache shareable and is stored in SGA memory

Database Link Calls
DO NOT USE USE

DECLARE
V_customer_name VARCHAR2(32); CREATE MATERIALIZED VIEW customers_snapshot
BEGIN ...
... AS
FOR c1 IN (SELECT …) SELECT customer_name .. FROM customers@remotedb;
LOOP
...
SELECT customer_name DECLARE
INTO v_customer_name V_customer_name VARCHAR2(32);
FROM customers@remotedb BEGIN
WHERE account_id = c1.account_id; ...
... FOR c1 IN (SELECT …)
END LOOP
/ ...
SELECT customer_name
INTO v_customer_name
FROM customers_snapshot
WHERE account_id = c1.account_id;
...
END
/

Excessive database link-based calls can affect application performance. Accessing a remote table or modifying a remote table over a database link within a
loop is not a scalable approach. For each access to a remote table, several SQL*Net packets are exchanged between the databases involved in the
database link.

Excessive Parsing
DO NOT USE USE

DECLARE DECLARE
... ...
BEGIN BEGIN
FOR c1_rec IN c1 FOR c1_rec IN c1
LOOP LOOP
-- Query customer details -- Query customer details
EXECUTE IMMEDIATE
'SELECT cust_first_name, cust_last_name, country_id
FROM customers SELECT cust_first_name, cust_last_name, country_id into
WHERE cust_id= '|| c1_rec.cust_id INTO l_cust_first_name, l_cust_first_name,
l_cust_last_name, l_cust_last_name,
l_country_id; l_country_id
... WHERE cust_id=c1_rec.cust_id;
END LOOP;
COMMIT; ...
END; END LOOP;
/ COMMIT;
END;
/

Do not use literals , Use Bind variables in SQL statements. Using literals causes excessive hard parsing stresses the library cache, thereby reducing the
application’s scalability and concurrency.

Excessive Commits
DO NOT USE USE
DECLARE
DECLARE commit_number pls_integer;
BEGIN
BEGIN
FOR c1_rec IN c1 commit_number := 0;
LOOP
FOR c1_rec IN c1
LOOP
SELECT cust_first_name, cust_last_name, country_id into
SELECT cust_first_name, cust_last_name, country_id into
l_cust_first_name, l_cust_first_name,
l_cust_last_name, l_cust_last_name,
l_country_id l_country_id
WHERE cust_id=c1_rec.cust_id;
WHERE cust_id=c1_rec.cust_id;
INSERT ...
INSERT ... UPDATE ...
UPDATE ... if commit_number > 10000
Commit; Then
END LOOP; commit;
commit_number := 0;
end if;
END;
/ commit_number := commit_number + 1;

END LOOP;

Commit;

END;
/

Frequent commits generate more redo, require Log Writer to flush the contents of log buffer to log file frequently, can lead to data integrity issues, and
consume more resources.

Type Usage in Spec Definetions
DO NOT USE USE

CREATE OR REPLACE PACKAGE check_bilgi CREATE OR REPLACE PACKAGE check_bilgi
IS IS
... ...
FUNCTION checkirtibatbilgi ( FUNCTION checkirtibatbilgi (
pcif IN t_mus_irtibat.client_no%TYPE, pcif IN NUMBER,
pirtbilgi IN t_mus_irtibat.irtbilgi%TYPE, pirtbilgi IN VARCHAR2,
pbilgitip IN t_mus_irtibat.bilgitip%TYPE, pbilgitip IN VARCHAR2,
pserino IN t_mus_irtibat.musserino%TYPE DEFAULT NULL pserino IN NUMBER DEFAULT NULL
) )
RETURN BOOLEAN; RETURN BOOLEAN;

.... ....

END; END;
/ /

If you are using packages in distributed environment ; do not use Type in Packege Spec definition. By this way, you can reduce database dependencies .

PL/SQL New Features
SIMPLE_INTEGER Datatype
The SIMPLE_INTEGER datatype is a subtype of the PLS_INTEGER datatype and can dramatical increase the speed of integer arithmetic

CREATE OR REPLACE PROCEDURE simple_integer_test_proc AS
l_start NUMBER;
l_loops NUMBER := 10000000;
l_pls_integer PLS_INTEGER := 0;
l_pls_integer_incr PLS_INTEGER := 1;
l_simple_integer SIMPLE_INTEGER := 0;
l_simple_integer_incr SIMPLE_INTEGER := 1;
BEGIN
l_start := DBMS_UTILITY.get_time;
FOR i IN 1 .. l_loops LOOP
l_pls_integer := l_pls_integer + l_pls_integer_incr;
END LOOP;
DBMS_OUTPUT.put_line('PLS_INTEGER: ' || (DBMS_UTILITY.get_time - l_start) || ' hsecs');

l_simple_integer := l_simple_integer + l_simple_integer_incr;
END LOOP;

DBMS_OUTPUT.put_line('SIMPLE_INTEGER: ' || (DBMS_UTILITY.get_time - l_start) || ' hsecs');
END simple_integer_test_proc;
/

PL/SQL New Features
Continue Statement
The CONTINUE statement jumps out of the current loop interation and starts the next one. It can be used on its own, or as part of CONTINUE WHEN statement

DECLARE
l_number NUMBER := 0;
BEGIN
FOR i IN 1 .. 100 LOOP
CONTINUE WHEN MOD(i,2) = 0;
-- Do something here!
l_number := l_number + 1;
END LOOP;

DBMS_OUTPUT.put_line('CONTINUE WHEN : ' || l_number);

l_number := 0;

FOR i IN 1 .. 100 LOOP
IF MOD(i,2) = 0 THEN
CONTINUE;
END IF;
-- Do something here!
l_number := l_number + 1;
END LOOP;

DBMS_OUTPUT.put_line('IF .. CONTINUE: ' || l_number);
END;
/

PL/SQL New Features
Sequences in PL/SQL expressions
The NEXTVAL and CURRVAL sequence pseudocolumns can now be accessed in PL/SQL expressions as well as queries.
CREATE SEQUENCE test1_seq START WITH 1000000;

DECLARE
l_start NUMBER;
l_value NUMBER;
BEGIN

l_value := test1_seq.NEXTVAL;
END LOOP;

SELECT test1_seq.NEXTVAL
INTO l_value
FROM dual;
END LOOP;
l_value := test1_seq.NEXTVAL;
END LOOP;

SELECT test1_seq.CURRVAL
INTO l_value
FROM dual;
END LOOP;

l_value := test1_seq.CURRVAL;
END LOOP;
END;

PL/SQL New Features
Dynamic SQL Enhancements
Native dynamic SQL and the DBMS_SQL package now support dynamic SQL statements larger than 32 KB.
The EXECUTE IMMEDIATE statement, OPEN-FOR statement and DBMS_SQL.PARSE procedure all accept SQL statements in the form of
CLOBs.
The DBMS_SQL.TO_REFCURSOR function converts a DBMS_SQL cursor ID into a REF CURSOR.

DECLARE
l_ref_cursor SYS_REFCURSOR;
l_cursor NUMBER;
l_count NUMBER := 0;
BEGIN
OPEN l_ref_cursor FOR 'SELECT * FROM emp';

l_cursor := DBMS_SQL.to_cursor_number(l_ref_cursor);

WHILE DBMS_SQL.fetch_rows(l_cursor) > 0 LOOP
l_count := l_count + 1;
END LOOP;
DBMS_OUTPUT.put_line('Employee Count: ' || l_count);

DBMS_SQL.close_cursor(l_cursor);
END;
/

PL/SQL New Features
Automatic Subprogram Inlining
Automatic subprogram inlining replace the subprogram calls with a copy of the code in the subprogram at compile time.
Reduce the overheads associated with calling subprograms

ALTER SESSION SET PLSQL_OPTIMIZE_LEVEL=2;

DECLARE
l_start NUMBER;
l_return NUMBER;

FUNCTION add_numbers (p_1 IN NUMBER,
p_2 IN NUMBER)
RETURN NUMBER AS
BEGIN
RETURN p_1 + p_2;
END add_numbers;

BEGIN

PRAGMA INLINE (add_numbers, 'YES');
l_return := add_numbers(1, i);
END LOOP;

END;
/

PL/SQL New Features
Native Compilation

In Oracle 11g, PL/SQL native compilation requires no C compiler, no DBA intervention and is fully supported in a RAC environment.

By setting the PLSQL_CODE_TYPE to a value of NATIVE, rather than the default value of INTERPRETED, code is compiled directly to machine code
and stored in the SYSTEM tablespace.

When the code is called, it is loaded into shared memory, making it accessible for all sessions in that instance. The %_PLSQL_OBJECT_SETTINGS
views include the current PLSQL_CODE_TYPE setting for each PL/SQL object.

Determining Whether to Use PL/SQL Native Compilation ( from Oracle® Database PL/SQL Language Reference)

While you are debugging program units and recompiling them frequently, interpreted mode has these advantages:
• You can use PL/SQL debugging tools on program units compiled for interpreted mode (but not for those compiled for native mode).
• Compiling for interpreted mode is faster than compiling for native mode.

After the debugging phase of development, in determining whether to compile a PL/SQL unit for native mode, consider:
• PL/SQL native compilation provides the greatest performance gains for computation-intensive procedural operations. Examples are data
warehouse applications and applications with extensive server-side transformations of data for display.
• PL/SQL native compilation provides the least performance gains for PL/SQL subprograms that spend most of their time running SQL.
• When many program units (typically over 15,000) are compiled for native execution, and are simultaneously active, the large amount of
shared memory required might affect system performance.

New Featues for Developers & DBAs
READ ONLY Tables

• In previous Oracle releases, tables could be made to appear read-only to other users by only granting the SELECT object privilege to them,
but the tables remained read-write for the owner.

• Oracle 11g allows tables to be marked as read-only using the ALTER TABLE command.

ALTER TABLE table_name READ ONLY;
ALTER TABLE table_name READ WRITE;

INVISIBLE indexes

• An invisible index is invisible to the optimizer as default.

• Using this feature we can test a new index without effecting the execution plans of the existing sql statements or we can test the effect of
dropping an index without dropping it.

• They are ignored by the optimizer unless the OPTIMIZER_USE_INVISIBLE_INDEXES parameter is set to TRUE at the instance or session level.

CREATE INDEX index_name ON table_name(column_name) INVISIBLE;

ALTER INDEX index_name INVISIBLE;

ALTER INDEX index_name VISIBLE;

SQL_monitoring

Oracle 11g automatically monitors SQL statements if they are run in parallel, or consume 5 or more seconds of CPU or I/O in a single execution.
This allows resource intensive SQL to be monitored as it is executing, as well as giving access to detailed information about queries once they are
complete.

New Features for Developers & DBAs
History Tables & Enterprise Manager
With history tables DBA can query historical performance data to solve problems . Enterprise Manager can help DBAs to probe problems.
Some important history tables

• DBA_HIST_ACTIVE_SESS_HISTORY
• DBA_HIST_SQLSTAT
• DBA_HIST_SQLTEXT
• DBA_HIST_SQL_PLAN
• DBA_HIST_SYSSTAT
• DBA_HIST_WAITSTAT

PL/SQL New Features
References

http://docs.oracle.com/cd/E11882_01/appdev.112/e25519/toc.htm

http://docs.oracle.com/cd/E11882_01/appdev.112/e25519/tuning.htm#CHDJJAGH

Oracle 11g PL/SQL notes

Recommended

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Oracle 11g PL/SQL notes (20)

Recently uploaded (20)

Oracle 11g PL/SQL notes