This transaction contains all check and monitoring functions for an application installation. At a particular point in time (checkpoint) which is either reached automatically or triggered interactively, the Watchdog can execute a subset of check rules from a given list of different check rules (topics). The following predefined checkpoints are available:

• Periodic (is executed, for example, every 30 minutes)
• Evening (typically used to monitor the daily closing of transactions not yet processed and pending messages)
• Morning

It is possible to program and run additional check rules for specific customers. All checkpoint results are generated in a consolidated document. Using the buttons on the first panel, it is possible to generate and immediately display the documents (also for individual check rules).

There are three different view types for each check rule:

• Counter (just one text line, that, where appropriate, contains Return Value and Threshold, or, for example, the text lines: 'All managers running' or 'No erroneous WFE'.)
• List (short display, typically the list of errors)
• Details (detailed display)

'Skip' means that the check rule is not run for this particular checkpoint.

A maximum return value (threshold) can be defined for each check rule defined as 'Periodic' (also 0). If this threshold is exceeded, it will be interpreted as an 'Error'. Such an error status triggers the following actions.

• Initially the error is posted as a procedure. The name of the procedure is composed of the name of the manager, service or report. If this is available, the procedure can then check other secondary conditions and, where applicable, can reset the error status. This then avoids having error messages for services that never run at certain times.
• A notify message is sent to the application administrator(s)
• The 'Notebook' in the office menu is marked red

If a checkpoint is exceeded without the Watchdog having been executed, the notebook in the office menu is also marked red.

Starting the Manager, if not available

If the switch is activated, at each periodic check, the Watchdog checks whether all the managers required are up and running. If a manager is, or several managers are, missing, the Watchdog carries out the following actions for the manager:

• If the 'old' manager process is still available, this is removed with a kill command
• The manager is re-started

The list of required managers is based on the configured list of managers as visible in the background manager controlling transaction.

Control of the interactive and background process window

These switches allow you to release and block the whole system at specified times. This allows you to enter different times for interactive foreground and background processes.

Operating times

Starting the system (interactive users)
- System is available for login

Starting the system (background processes)
- Managers will be started

Stopping the systems (interactive users):
- First of all the system is blocked so that no other users can be logged in.
- Afterwards, all interactive process that are still running will be closed.

Stopping the system (background processes)
- The managers will be shut down. In addition, a time can be entered when a Notify message is sent to the interactive users indicating that the system will be shut down in x minutes.

All times apply to the time zone set up for the Watchdog.

The Watchdog also provides a function to immediately shut down and block the system. This can be launched by the Control Background Manager (SYSMGR) transaction.

The watchdog provides the possibility to hand over the result to an external auditing system.
If external monitoring is turned on, the results of the checks, where external monitoring is checked, are written to a file named in 'Filename'.
This file can be sent to another location with a script defined in 'Jobname'.

Turning on 'Generate volume report' will write the result of every check (morning check, evening check and periodical check) into the database table SYI.
Every single check will generate one row in this table.

Managers running in foreground mode always need to be started manually.

Watchdog
Configuration
Show Details
Show Processes
Email
Email technical configuration
External monitoring

This panel is used to specify Watchdog configuration settings. The following table shows all available options, listing designated check rules and typical configurations for checkpoints. The columns in bold letters are displayed in the application. Manually changeable columns are marked correspondingly in the header.

Note: This function is more comprehensive on Unix clients, as additional check rules are available.

HTM and PDF are available output formats for statistics. In the 'Max. # of entries per list' field, limits for messages displayed per list is defined.

(*) = Currently only available for operating system AIX

This panel is equivalent to the 'Watchdog Info' panel (SYSWDI). The details of the last checks can be launched here. Furthermore, checks (periodic, morning, evening) can be started. Date and time of the last performed check are also displayed.

The results of the last performed checks can be displayed by clicking [Show Results]. The checks can be run again by clicking [Execute Check].

The following buttons are available for more results:

• [List] - Creating and displaying the relevant report
• [Details] - Creating and displaying the relevant details
• [Info] - Starting the associated info system
• [Manage] - Starting the associated manage transactions

Note: If not all functions are available, (particularly 'Manage' functions), the standard profile functionality has to be modified.