System Performance

As a system administrator, you make sure that your system is configured to guarantee the best performance possible.

Your Server computer is the core part of AE and data-center automation. Therefore, using adequate performance and security features is crucial. For more information, see Automic Automation System Requirements and Sizing.

Important!

  • Deactivate low-current functions and dynamical cycle adjustment in the BIOS on the Automation Engine computer if x core AMD CPU is used.

  • When writing a trace, the trace file must be open and closed for each job, which can lead to significant performance losses. Therefore, make sure traces are activated only when necessary.

Performance Checks

System administrators can define how the system collects and stores performance data, which facilitates a fast and precise analysis of the system. The parameters relevant for this performance checks are stored in the UC_SYSTEM_SETTINGS variable and allow you to define different aspects such as if the system should gather performance data, how often, how long the data is kept, and so on. For more information, see PERFORMANCE Parameters.

The AE REST API also provides an endpoint for this purpose. For more information, see AE REST API - AE Performance Metrics.

The system also monitors its performance metrics and checks for performance incidents as defined in the UC_PERFORMANCE_THRESHOLDS variable. For more information, see UC_PERFORMANCE_THRESHOLDS - Performance Check.

This function surfaces possible performance issues early and gives administrators the chance to deal with possible issues themselves and reduce possible outages or, at least, reduce the severity of a potential performance issue.

Server processes measure the system's performance. For example, every hour, a Work Process (WP) picked at random measures the database performance. Other metrics are measured by different server processes in the interval defined in the PERFORMANCE_MEASURE_INTERVAL key of the UC_SYSTEM_SETTINGS, see PERFORMANCE Parameters.

The check takes only a few seconds, depending on the database type. The JWP with the PER Role analyzes the data and the information about possible issues is written to that JWP log and kept in a performance table in the AE database. For more information, see JWP Roles.

For example:

U00005901 The following metrics were classified as suspicious:
U00005902 AUTOMIC#WP001 ([IP OR DNS NAME]) at UTC 2024-01-15 00:00:10.0: {db-perf=190.0,rollback=1244370.0,select=2623732.0,
fetch=6656.0,bindpar=3204.0,insert=1328542.0,close=15371.0,cpu-perf=9.2744795E7,bindcol=5838.0}

The performance check adds all the values and displays them in milliseconds. If the resulting value surpasses the thresholds that have been defined in the UC_PERFORMANCE_THRESHOLDS variable, a corresponding warning is displayed in the Automic Web Interface, namely:

  • "System might not meet minimum requirements. Please check your infrastructure (Automation Engine, database, network). for the system requirements

  • "Message queue threshold exceeded: <queue1, queue2>" for message queues

If one or both of these messages appear, contact the database or system administrator so that the system's performance can be optimized. Once the performance improves, the warning disappears automatically after the next measurement.

Important! The performance check can be deactivated using the PERFORMANCE_ACTIVE key of the performance parameters in the UC_SYSTEM_SETTINGS, see PERFORMANCE Parameters. However, it can only be deactivated using the for the message queues, not for the database as these measurements are crucial for the system's performance. The information provided is indicative of issues outside Automic Automation on any layer, such as the JDBC, the I/O on the AE or database server, the network between the AE and database servers, and so on. Therefore, solving a performance issue might require a cross-team effort (DB, system, server administrators, network team, and so on). As the person or team in charge, make sure you mitigate the issue by optimizing the system, not by changing the thresholds defined in the variable. However, if necessary, they can be adapted.

Optionally, you can use the Database Metric Visualization tool to get your performance metrics displayed in charts, which can help when troubleshooting possible performance issues. For more information, see Database Metric Visualization.