Configuring Adaptive Thresholds

Figure 2.1: Enabling Metric Baselines for the first time.


Figure 2.2: Creating a Metric Baseline using a rolling time period.


Figure 2.3: Failure during creation of a static Metric Baseline (i.e., insufficient statistics).


Figure 2.4: Successful creation of a static Metric Baseline (i.e., sufficient statistics).


Figure 2.5: Choosing a static Metric Baseline for enforcing Adaptive Thresholds.


Figure 2.6: Selecting which Adaptive Thresholds metrics to manage.


Figure 2.7: Setting alert threshold values for the selected Adaptive Thresholds metrics.


Figure 2.8: Viewing a graph of the significance of Adaptive Thresholds metrics depending on the time series selected.


Figure 2.9: Picking the nine Adaptive Thresholds metrics that will be displayed on the Baseline Normalized Metrics page.


Figure 2.10: Displaying the chosen nine Adaptive Threshold metrics and any recent "deflections" from the specified significance factors.


Figure 2.11: Testing the Adaptive Threshold for Number of Transactions.


Figure 2.12: Disabling Metric Baselines once they've been enabled.


 
 
----- 
-- Listing 2.1: Shell script to "overtask" an Oracle 10gR2 database
--              to force detection by Adaptive Thresholds
-----

/*
|| Create table for DML statement execution to simulate "overtasking"
*/
DROP TABLE hr.cranktrx PURGE;
CREATE TABLE hr.cranktrx (
    c NUMBER
);

/*
|| PL/SQL anonymous block to ramp up number of transactions / second
*/

SET ECHO OFF
SET FEEDBACK OFF
SET VERIFY OFF

DECLARE
    tps NUMBER := &1;
BEGIN
    tps := (tps / 2);
    FOR i IN 1..180
	LOOP
    FOR i IN 1..tps 
        LOOP
            INSERT INTO hr.cranktrx VALUES(1);
            COMMIT;
            DELETE hr.cranktrx WHERE rownum < 2;
            COMMIT;
        END LOOP;
    DBMS_LOCK.SLEEP(1);
END LOOP;
END;
/