Logs and alerts
During run time, Pega Platform™ captures exceptions, warnings, and debugging information in log files. These logs provide essential data for diagnosing system issues.
In Pega Platform, logs and alerts go beyond basic diagnostics. They support proactive monitoring, resilience planning, and security enforcement. Lead System Architects (LSAs) must configure, interpret, and act on these signals to maintain system performance and stability.
LSAs are responsible for maintaining system health, resolving performance bottlenecks, and ensuring security compliance. Logs and alerts help triage issues, identify root causes, and communicate findings across teams.
Logs
Pega logs are organized into categories to help you monitor system health and troubleshoot issues. Developers use logs during development to record informational messages, inspect property values, or explain why a step failed. Logs are especially useful when tools such as the Clipboard or Tracer are unavailable.
Log files support debugging and capture data about internal operations, system performance, and security. You can turn logging on or off as needed.
Log levels
Pega Platform supports multiple log levels to indicate the severity of events: FATAL, ERROR, ALERT, WARN, INFO, and DEBUG. For example, if the log level is set to ALERT, the system records messages with ALERT, ERROR, and FATAL severity. Using the appropriate log level is important because it affects the size of the log files. There is also an ALL level that logs every message. Each log category has an automatic reset interval, and you can reset the log level when detailed logging is no longer required.
For more information, see Changing the log level of a single log category and Resetting the log levels of all log categories.
Rolling log files
To generate a new log file every day, you update the prlog4j2.xml file to generate a new log file at the start of each day or on a periodic basis, rather than just at system startup. This process is called rolling the log file. For more information, see Rolling log file.
Debugging issues related to autopopulated properties can be difficult without metadata. Creating logs with metadata can help with troubleshooting autopopulated properties. pxAutoPopulate is the log category for logging autopopulated properties. For more information about generating logs for more information about autopopulated properties and related settings, see Generating logs for autopopulated properties.
Alerts
Alerts in Pega Platform are generated when performance or security issues occur. The system records alerts in log files for analysis.
Performance alerts are triggered when a Rule runs longer than the defined threshold. Alerts prefixed with PEGA indicate performance issues. Alerts that begin with SECU indicate security concerns. Some performance alerts might also relate to database delays. For example:
- PEGA0001 is triggered when the elapsed time for an HTTP interaction exceeds the configured threshold. This alert might result from long-running calculations, database connection delays, or slow responses from external services.
- PEGA0020 is generated when the time taken to interact with an external system or database exceeds the threshold. This alert often indicates network issues affecting existing or newly created connections.
Each alert follows a structured format to aid understanding. For example, PEGA0011 is raised when a service operation or parse rule takes longer than expected. It monitors five key metrics:
- pxTotalReqTime: Total request time (alerts typically appear for durations over one second).
- pxServiceImpMapReqTime: Time taken for request mapping.
- pxServiceOutMapReqTime: Time taken for response mapping.
- pxServiceActivityTime: Time taken to run the service activity.
- pxServiceParseRuleTime: Time taken to run Parse XML, Parse Delimited, or Parse Structured Rules.
For more information, see Alerts overview.
Advanced monitoring and alerting
In today's high-availability enterprise environments, reactive troubleshooting is no longer sufficient. LSAs must adopt proactive monitoring strategies that anticipate issues before they disrupt operations. Pega Platform supports this approach with intelligent alerting, dynamic threshold management, and early detection features that turn logs and alerts into actionable insights.
Automated notification configuration
Pega Diagnostic Center (PDC) now offers comprehensive notification features that can automatically alert stakeholders when critical events occur. Notifications can undergo configuration based on:
- Severity level
- Affected component
- Business impact
Set up notifications for memory threshold breaches, CPU utilization spikes, integration failures, and security events.
Alert threshold management
Effective monitoring requires carefully calibrated thresholds that balance sensitivity with noise reduction. The enhanced PDC interface enables LSAs to:
- Set dynamic thresholds based on historical performance data.
- Adjust thresholds for peak versus off-peak hours.
- Adapt thresholds for special event periods.
Early detection strategies
Proactive monitoring involves identifying patterns that precede system issues. LSAs must monitor:
- Queue growth rates
- Response time trends
- Memory allocation patterns
- Integration response times
By establishing baseline performance metrics and detecting deviations early, LSAs can prevent minor issues from escalating into system-wide problems.
Check your knowledge with the following interaction: