Issue identification and research with Pega Predictive Diagnostic Cloud
Pega Predictive Diagnostic Cloud™ (PDC) helps you identify areas that you need to work on to improve your system's performance. Use Improvement Plan to view a summary of issues that affect the performance of your monitored system. Use Event Viewer to conduct an in-depth investigation.
- Improvement Plan features for identifying and researching issues in your system
- Event Viewer features for identifying and researching issues in your system
To ensure the smooth functioning of your system, use Improvement Plan to identify issues that require immediate attention:
- View cases prioritized by their urgency for each affected area of your system. Investigate the top items in all the categories.
- Check whether current top cases are still affecting your system significantly.
- Monitor case urgency trends to ensure that issues do not escalate further.
- Assign cases to the appropriate problem solvers.
The performance statistics in Improvement Plan are divided into categories that represent the following areas:
- Application Logic – Issues related to the configuration that is used to incorporate the business logic in your application.
- Connectors – Issues related to the connectors that are used in your application.
- Database – Issues that occur in your database configuration.
- Decisioning – Issues related to Decision Strategy Manager (DSM).
- Exceptions – Issues related to errors in the code.
- Operations – Issues related to the run-time operation of your application.
- Pega Platform Tuning – Issues related to your Pega Platform configuration.
You can display up to 500 cases in a category, sorted by urgency. To find the cases that you want to investigate more quickly, you can sort and filter cases in a category alphabetically by case, assignment, and the last time when PDC detected an associated event.
You can view the detailed description, root cause analysis, and recommended actions for each case.
The following columns provide a quick view into the cases:
- Assigned – The user or workbasket to which the case is assigned.
- Affected – The process area in your system that is affected by an issue: web interactions, service interactions, or batch processing.
- Impact – The amount of time that your system spent on an issue in the last seven days, and a percentage increase or decrease compared to the previous week.
- Occurrences – The frequency with which this issue occurred within the last seven days and a percentage increase or decrease compared with the previous week.
- Trend – The red (higher) and green (lower) arrows show whether a case is gaining or decreasing in importance.
- Last seen – The last occurrence of an associated event.
You can configure notifications so that you periodically receive emails with the detailed Improvement Plan or emails with a summary of each category. For more information, see Managing PDC notification subscriptions.
In Event Viewer, you can access a live feed of diagnostic data from all of the nodes in your system.
Use Event Viewer to perform an in-depth checkup:
- Check whether your system has encountered any exceptions, which are unplanned events that interrupt the flow of a running program.
- Investigate events that preceded system failure.
- Evaluate the circumstances of an issue:
- Investigate events that coincided with diminished performance to find areas for improvement.
- Assess whether a node that has affected performance is running correctly.
You can view data from up to the last 14 days to analyze issues that occurred during that time. For example, if you know when your system failed, you can view the alerts, exceptions, and PAL events that were generated at that time. By analyzing these events, you can determine what caused the selected system to fail.
For more information about the event types in PDC, see Cases in Pega Predictive Diagnostic Cloud.
To identify the main source of events that occurred on the selected system in the specified time frame, you can group events by the following criteria:
- Case ID – Show events related to a case as one item. This grouping helps you find the cases that most affect your system's performance.
- Event Type – Show events of the same type as one item. This grouping helps you find the type of event that occurs most frequently.
- Application – Show events related to the same application as one item. This grouping helps you find the application that generates the most events.
- Operator – Show events that a single user's actions trigger as one item. This grouping helps you find the users that cause the most performance issues with your system.
- Node – Show events that occurred on the same node as one item. This grouping helps you find the slowest nodes in your system.
- Requestor – Group events that share the same requestor. This grouping helps you analyze which requestors generated the most events.
In the Event Viewer, you can filter all events by the following criteria:
- Date – Select the date to view more details about an event.
- Type – Select the type of event to see: exception, alert, or PAL event.
A special icon in the Type column informs you about urgent events that require your immediate attention.
- Event – Select the event title or exception message.
- Case ID – Select a specific case to see all related events.
- Application – Select an application to view.
- Operator – Select a user account to view the issues that a specific user was experiencing.
- Requestor – Select the session to search for events.
- Int ID – Select an interaction ID to view.
- Node – Select the system from which the events came.
- Maintenance mode – View only events that occurred when maintenance mode was on.
- Requestor type – Select one or more of the following requestor types: Web, Service, Background, and Async.
- Interval – Click Date and time filters to filter by the time period for the data. If the time that you need is not available in the list, select Custom time, and then configure the date and time for when the events that you want to view were generated.