Oracle 12c DBMS_ADDM Demos

Oracle DBMS_ADDM Version 12.1.0.1

General Information

Library Note

Morgan's Library Page Footer

The Library is currently in the process of being upgraded from Oracle Database Version 11.2.0.3 to 12.1.0.1. Demos are being upgraded to reflect the new Container paradigm as well as EBR (Edition Based Redefinition) and may contain references to CDBs, PDBs, and other objects you may not be familiar with such as CDB_OBJECTS_AE: Welcome to 12c.

Purpose

Create and execute an tasks to analyze all instances of a database.

AUTHID

CURRENT_USER

Dependencies

DBMS_ADVISOR	DBMS_UTILITY	PRVT_ADVISOR
DBMS_MANAGEMENT_PACKS	DBMS_WORKLOAD_REPLAY	PRVT_HDM
DBMS_REPORT	DUAL	USER_ADVISOR_TASKS
DBMS_SYS_ERROR

Documented

Partially

First Available

11.1.0.6

Security Model

Owned by SYS with EXECUTE granted to PUBLIC

Source

{ORACLE_HOME}/rdbms/admin/dbmsaddm.sql

Subprograms

ANALYZE_DB	DELETE	GET_REPORT
ANALYZE_INST	DELETE_FINDING_DIRECTIVE	INSERT_FINDING_DIRECTIVE
ANALYZE_PARTIAL	DELETE_PARAMETER_DIRECTIVE	INSERT_PARAMETER_DIRECTIVE
COMPARE_CAPTURE_REPLAY_REPORT	DELETE_SEGMENT_DIRECTIVE	INSERT_SEGMENT_DIRECTIVE
COMPARE_DATABASES	DELETE_SQL_DIRECTIVE	INSERT_SQL_DIRECTIVE
COMPARE_INSTANCES	GET_ASH_QUERY	REAL_TIME_ADDM_REPORT
COMPARE_REPLAY_REPLAY_REPORT

ANALYZE_DB

Create and execute an ADDM task to analyze all instances of a database for a range of snapshots

dbms_addm.analyze_db(
task_name      IN OUT VARCHAR2,
begin_snapshot IN     NUMBER,
end_snapshot   IN     NUMBER,
db_id          IN     NUMBER := NULL);

SELECT dbid FROM v$database;

set pagesize 25
set linesize 121
col instart_fmt format a20
break on instart_fmt
col inst_name format a12 heading 'Instance';
col db_name format a12 heading 'DB Name';
col snap_id format 99999990 heading 'Snap Id';
col snapdat format a18 heading 'Snap Started' just c;
col lvl format 99 heading 'Snap|Level';
set heading on;
break on inst_name on db_name on host on instart_fmt skip 1;
ttitle off;

SELECT TO_CHAR(s.startup_time,' DD MON "at" HH24:MI:SS') INSTART_FMT,
di.instance_name INST_NAME, di.db_name DB_NAME, s.snap_id SNAP_ID,
TO_CHAR(s.end_interval_time,'DD MON YYYY HH24:MI') SNAPDAT,
s.snap_level LVL
FROM dba_hist_snapshot s, dba_hist_database_instance di
WHERE di.dbid = s.dbid
AND di.instance_number = s.instance_number
AND di.startup_time = s.startup_time
ORDER BY snap_id;

set serveroutput on

DECLARE
tname VARCHAR2(30) := 'UW_TASK1';
BEGIN
dbms_addm.analyze_db(tname, 2437, 2504, 1701481905);
dbms_output.put_line(tname);
END;
/

col execution_name format a15
col finding_name format a30
col parent format 99
col impact format 99999999999
col message format a60
col more_info format a60

SELECT execution_name, finding_name, parent, impact
FROM dba_addm_findings
WHERE task_name = 'UW_TASK1';

SELECT execution_name, finding_name, message
FROM dba_addm_findings
WHERE task_name = 'UW_TASK1';

SELECT execution_name, finding_name, more_info
FROM dba_addm_findings
WHERE task_name = 'UW_TASK1';

ANALYZE_INST

Create and execute an ADDM task to analyze a specific instance of a database for a range of snapshots

dbms_addm.analyze_inst(
task_name       IN OUT VARCHAR2,
begin_snapshot IN     NUMBER,
end_snapshot    IN     NUMBER,
instance_number IN     NUMBER := NULL,
db_id           IN     NUMBER := NULL);

SELECT dbid FROM v$database;

SELECT inst_id
FROM gv$instance;

set pagesize 25
set linesize 121
col instart_fmt format a20
break on instart_fmt
col inst_name format a12 heading 'Instance';
col db_name format a12 heading 'DB Name';
col snap_id format 99999990 heading 'Snap Id';
col snapdat format a18 heading 'Snap Started' just c;
col lvl format 99 heading 'Snap|Level';
set heading on;
break on inst_name on db_name on host on instart_fmt skip 1;
ttitle off;

SELECT TO_CHAR(s.startup_time,' DD MON "at" HH24:MI:SS') INSTART_FMT,
di.instance_name INST_NAME, di.db_name DB_NAME, s.snap_id SNAP_ID,
TO_CHAR(s.end_interval_time,'DD MON YYYY HH24:MI') SNAPDAT,
s.snap_level LVL
FROM dba_hist_snapshot s, dba_hist_database_instance di
WHERE di.dbid = s.dbid
AND di.instance_number = s.instance_number
AND di.startup_time = s.startup_time
ORDER BY snap_id;

set serveroutput on

DECLARE
tname VARCHAR2(30) := 'UW_TASK2';
BEGIN
dbms_addm.analyze_inst(tname, 2387, 2406, 1, 1701481905);
dbms_output.put_line(tname);
END;
/

desc dba_addm_findings

SELECT task_name
FROM dba_addm_findings;

SELECT task_name
FROM dba_addm_findings
WHERE task_name LIKE 'UW%';

col finding_name format a30
col impact format 99999.99

SELECT execution_name, finding_name, parent, impact, message, more_info
FROM dba_addm_findings
WHERE task_name = 'UW_TASK2';

desc dba_addm_tasks

SELECT DISTINCT task_name, advisor_name
FROM dba_addm_tasks;

col advisor_name format a15
col how_created format a15

SELECT advisor_name, status, how_created, system_task, requested_analysis, actual_analysis
FROM dba_addm_tasks
WHERE task_name = 'UW_TASK2';

ANALYZE_PARTIAL

Create and execute an ADDM task to analyze a partial list of instances for a range of snapshots

dbms_addm.analyze_partial(
task_name        IN OUT VARCHAR2,
instance_numbers IN     VARCHAR2,
begin_snapshot   IN     NUMBER,
end_snapshot     IN     NUMBER,
db_id            IN     NUMBER := NULL);

COMPARE_CAPTURE_REPLAY_REPORT (new 12.1)

Create a report comparing the performance of a capture to a replay

dbms_addm.compare_capture_replay_report(
replay_id IN NUMBER,
report_type IN VARCHAR2 := 'HTML')
RETURN CLOB;

TBD

COMPARE_DATABASES (new 12.1)

Create a report comparing the performance of a database over two different time periods or the performance of two different databases over two different time periods

dbms_addm.compar_databases(
base_dbid          IN NUMBER,
base_begin_snap_id IN NUMBER,
base_end_snap_id   IN NUMBER,
comp_dbid          IN NUMBER,
comp_begin_snap_id IN NUMBER,
comp_end_snap_id   IN NUMBER,
report_type        IN VARCHAR2 := 'HTML')
RETURN CLOB;

TBD

COMPARE_INSTANCES (new 12.1)

Create a report comparing the performance of a single instance over two different
time periods or the performance of two different instances over two different time periods

dbms_addm.compare_instances(
base_dbid          IN NUMBER,
base_instance_id   IN NUMBER,
base_begin_snap_id IN NUMBER,
base_end_snap_id   IN NUMBER,
comp_dbid          IN NUMBER,
comp_instance_id   IN NUMBER,
comp_begin_snap_id IN NUMBER,
comp_end_snap_id   IN NUMBER,
report_type        IN VARCHAR2 := 'HTML')
RETURN CLOB;

TBD

COMPARE_REPLAY_REPLAY_REPORT (new 12.1)

Create a report comparing the the performance of two replays

dbms_addm.compare_replay_replay_report(
replay_id1 IN NUMBER,
replay_id2 IN NUMBER,
report_type IN VARCHAR2 := 'HTML')
RETURN CLOB

TBD

DELETE

Delete an ADDM task

dbms_addm.delete(task_name IN VARCHAR2);

exec dbms_addm.delete('UW_TASK1');

DELETE_FINDING_DIRECTIVE

Delete an ADDM directive

dbms_addm.delete_finding_directive(task_name IN VARCHAR2, dir_name IN VARCHAR2);

SELECT task_name, directive_name, description
FROM dba_addm_task_directives;

DELETE_PARAMETER_DIRECTIVE

Delete an ADDM directive

dbms_addm.delete_parameter_directive(task_name IN VARCHAR2, dir_name IN VARCHAR2);

TBD

DELETE_SEGMENT_DIRECTIVE

Delete an ADDM directive

dbms_addm.delete_segment_directive(task_name IN VARCHAR2, dir_name IN VARCHAR2);

TBD

DELETE_SQL_DIRECTIVE

Delete an ADDM directive

dbms_addm.delete_sql_directive(task_name IN VARCHAR2, dir_name IN VARCHAR2);

TBD

GET_ASH_QUERY

Retrieves the SQL text of an ASH query for a finding

dbms_addm.get_ash_query(task_name IN VARCHAR2, finding_id IN NUMBER)
RETURN VARCHAR2;

-- first run the ANALYZE_DB demo above over a range that has "real" work

SELECT finding_id, finding_name, message
FROM dba_advisor_findings
WHERE task_name = 'UW_TASK1';

SELECT dbms_addm.get_ash_query('UW_TASK1', 1) FROM dual;

col task_name format a25
col message format a50

SELECT task_name, finding_id, finding_name, message
FROM dba_advisor_findings;

SELECT dbms_addm.get_ash_query('ADDM:2495529197_1_25', 1) FROM dual;
SELECT dbms_addm.get_ash_query('ADDM:2495529197_1_3', 1) FROM dual;

GET_REPORT

Returns the ADDM report

dbms_addm.get_report(task_name IN VARCHAR2) RETURN CLOB;

set long 1000000
set serveroutput on

DECLARE
rptlob CLOB;
BEGIN
rptlob := dbms_addm.get_report('UW_TASK1');
dbms_output.put_line(rptlob);
END;
/

SELECT * FROM TABLE(dbms_xplan.display_cursor('1tkztbrcvfcdw'));

INSERT_FINDING_DIRECTIVE

Insert a directive filtering a finding, either for a task or system wide

dbms_addm.insert_finding_directive(
task_name           IN VARCHAR2,
dir_name            IN VARCHAR2,
finding_name        IN VARCHAR2,
min_active_sessions IN NUMBER := 0,
min_perc_impact     IN NUMBER := 0);

TBD

INSERT_PARAMETER_DIRECTIVE

Insert a directive filtering a recommendations to change a system parameter

dbms_addm.insert_parameter_directive(
task_name IN VARCHAR2,
dir_name IN VARCHAR2,
parameter_name IN VARCHAR2);

TBD

INSERT_SEGMENT_DIRECTIVE

Insert a directive filtering recommendations to run the segment advisor

Overload 1

dbms_addm.insert_segment_directive(
task_name       IN VARCHAR2,
dir_name        IN VARCHAR2,
owner_name      IN VARCHAR2,
object_name     IN VARCHAR2 := NULL,
sub_object_name IN VARCHAR2 := NULL);

TBD

Overload 2

dbms_addm.insert_segment_directive(
task_name IN VARCHAR2,
dir_name IN VARCHAR2,
object_number IN NUMBER);

TBD

INSERT_SQL_DIRECTIVE

Insert a directive filtering SQL recommendations, either for a task or system wide

dbms_addm.inset_sql_directive(
task_name           IN VARCHAR2,
dir_name            IN VARCHAR2,
sql_id              IN VARCHAR2,
min_active_sessions IN NUMBER := 0,
min_response_time   IN NUMBER := 0);

TBD

REAL_TIME_ADDM_REPORT (new 12.1)

Returns a real-time ADDM report for the previous 5 minutes

dbms_addm.real_time_addm_report RETURN CLOB;

spool c:\temp\realtime_addm.html

SELECT dbms_addm.real_time_addm_report
FROM dual;

spool off

-- open the report in a browser