THRESHOLDS Policy Item

The Thresholds Definition panel is displayed if you select policy item THRESHOLDS on the Policy Selection panel for applications, as shown in Figure 1.

Figure 1. Thresholds Definition Panel

   COMMANDS  HELP                                                               
 ------------------------------------------------------------------------------ 
 AOFGDYNA                    Thresholds Definition                Line 00000001 
 Command ===>                                                  Scroll ===> CSR  
                                                                                
 Entry Type : Application           PolicyDB Name   : USER_PDB                  
 Entry Name : APPL1                 Enterprise Name : USER_ENTERPRISE           
                                                                                
 Resource   : APPL1                                                             
                                                                                
 Critical Number  . . . . 5          (1 to 50)                                  
 Critical Interval  . . . 00:10:00   (hh:mm:ss or hhmmss, 00:00:01 to 24:00:00) 
                                                                                
 Frequent Number  . . . . 3          (1 to 50)                                  
 Frequent Interval  . . . 01:00:00   (hh:mm:ss or hhmmss, 00:00:01 to 24:00:00) 
                                                                                
 Infrequent Number  . . . 3          (1 to 50)                                  
 Infrequent Interval  . . 12:00:00   (hh:mm:ss or hhmmss, 00:00:01 to 24:00:00)

The information provided in the THRESHOLDS dialogs is used to determine when an operator should be notified of an error situation. An attempt at recovering an error (such as an abend of a subsystem) is handled by automation, but an operator may want to know if this error is occurring often.

Thresholds can be set that let the operator know if certain errors are occurring infrequently, frequently or have reached a critical stage where the recovery process should be ended (to avoid endless loops). This is done by specifying how many times an error must happen in a certain time period for each error situation.

The entries on this panel can be inherited from a class, in which case they are displayed in a different color. Because all of the thresholds are inherited together, if you edit one of them, all of the thresholds change color to indicate that they are now particular to that instance because they are no longer inherited.

You can find out which class a value has been inherited from by typing the command classname (or its abbreviation cn) on the command line.

You can define three different threshold levels:

Infrequent: Issues a warning message and restart processing is done.
Frequent: Issues a warning message, but restart processing is still done.
Critical: Sets the resource to a status of BROKEN, stops the restart automation and issues a message.

You can enter data in the following fields:

Number: This field is used to define the number of occurrences of an error that creates the particular error level. The range of allowable numbers is 1 to 50.
Interval: The number of errors entered in the Number field must occur in a certain time period in order to create a specific error level. You specify this time period in the Interval field. You can specify it in the form HH:MM:SS or HHMMSS. Valid values are 00:00:01 to 24:00:00.

The values entered in the Number and Interval fields should be such that the threshold levels are reached in a logical manner (a critical threshold should not be reached before a frequent threshold).

Not all three threshold levels need be specified, for example, if an application needs only a specification for the critical number and interval, you can leave the frequent and infrequent fields empty.

For more information on setting thresholds, see the section How SA z/OS Uses Error Thresholds in IBM Z® System Automation User's Guide.