Monitoring event queue efficiency through Application Insights
Summarize
Summarized using AI
This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.
Summary of Monitoring event queue efficiency through Application Insights
ServiceNow customers can monitor event queue performance using Application Insights by analyzing the rate at which events are logged and processed. This monitoring helps detect bottlenecks, anomalies, and performance patterns in event handling to ensure system efficiency and reliability.
Show less
Key Features
- Events Graphs: Access the Events graphs via All > Application Insights > Application Insights > Overview to monitor incoming and processed event rates, detect anomalies, and compare totals for bottleneck identification.
- Bottleneck Detection: Identify issues when spikes in Events Logged are not matched by spikes in Events Processed, signaling processing delays or failures.
- Load Monitoring: Distinguish between high loads and potential loops causing continuous event streams by observing consistent spikes in both logged and processed events.
- Node-Level Analysis: Drill down into detailed graphs to analyze performance issues at the node level for targeted troubleshooting.
- Trend and Pattern Analysis: Use the 1-Day Moving Average on Events Processed and Events Logged graphs to identify recurring spikes or patterns over time, including weekly cycles.
- Correlation with System Activities: Overlay diagnostic events to correlate spikes with system changes, such as update set installations, aiding root cause analysis.
- Event Detail Investigation: Select data points at spike onsets to review events occurring five minutes before and after, facilitating precise issue investigation.
- Event Queue Metrics: Analyze individual event queues over selectable periods (1-day, 7-day, 30-day) using the Event Queue table, which provides metrics on logged, queued, unprocessed, processed events, and processing durations.
Key Outcomes
- Improved visibility into event queue performance and processing efficiency.
- Early detection of processing bottlenecks and anomalies to maintain system stability.
- Ability to correlate event performance issues with system activities for faster root cause identification.
- Enhanced capacity planning and resource allocation by understanding load patterns and queue performance.
- Detailed event queue metrics empower targeted performance optimization and troubleshooting.
You can monitor event queue performance in Application Insights by comparing and analyzing the rate at which events are logged and processed.
You can monitor the relationship between events logged and events processed by viewing the Events graph on the Overview tab.
Use the Events graphs to do the following:
- Monitor the rate of incoming events
- Monitor the rate of processed events
- Detect anomalies in processing events
You access the Events graphs by navigating to .
- Look for bottlenecks by comparing the Events Logged and Events Processed totals. A spike in the Events Logged number without a corresponding spike in the Events Processed number indicates a problem with processing events.
- A high and consistent spike in both the Events Logged and Events Processed figures indicates the system is receiving and processing a large load of events. Look for loops or conditions that might be causing a continuous stream of incoming events.
Dig deeper into potential performance issues by drilling down to analyze issues at the node-level in the detail graphs.
- Analyze the processing of events over time by comparing processed events to the 1-Day Moving Average shown on the Events Processed graph.
- Look for patterns in event logging over time. Determine whether the same spike is happening at the same time every week.
- Look for correlations between the count of events logged and system activities by overlaying diagnostic events on the graph. For example, if you see a spike in the Events Logged number and you notice that it coincides with the installation of an update set, you can investigate the update set to determine why it caused the spike in events logged.
- Investigate the cause of an issue by selecting a data point at the start of spike to view a list of events created 5 minutes before and 5 minutes after the issue.
- Analyze the rate of incoming events by comparing logged events to the 1-Day Moving Average total on the Events Logged graph.
- Identify potential performance issues by comparing metrics for
individual event queues in the Event Queue table.
- Focus on a 1-day, 7-day, or 30-day period by selecting a day range.
- Identify which queue had the most logged events by sorting on the Logged events in range column. An event is logged when it is inserted into the Events [sysevent] table.
- Identify which queue had the most queued events by sorting on the Queued events in range column. An event is queued when it is assigned to a specific event queue in the Events [sysevent] table.
- Identify which queue had the most unprocessed events by sorting on the Unprocessed events in range column.
- Identify which queue took the most amount of time processing events in the selected day range by sorting on the Processing duration in range column.
- Identify which queue processed the most events in the selected day range by sorting on the Processed events in range column.
- Identify which queue took the most amount of time processing events on average in the selected day range by sorting on the Average processing duration in range column.