Monitoring users and transaction performance through Application Insights
Summarize
Summary of Monitoring users and transaction performance through Application Insights
ServiceNow customers can maintain their instance’s health and performance by monitoring key user and transaction metrics using the Application Insights Users and Transactions graph. This tool enables detection of anomalies and performance deviations by analyzing the relationship between user activity, transaction load, and average response times. However, starting with the Zurich release, Application Insights is no longer supported or enhanced. Customers are encouraged to evaluate the Instance Observer product included with ServiceNow Impact packages and consult their Account team for guidance.
Show less
Key Features
- Users and Transactions graph: Accessed via All > Application Insights > Application Insights > Overview, it shows transaction performance, transaction load, and the number of logged-in users.
- Performance analysis: Compare Transaction Count and Logged-in Users to analyze system slowness; spikes in both may indicate false alarms.
- Impact assessment: Evaluate if high transaction loads affect performance by comparing Average Response Time with Transaction Count.
- Drill-down capabilities: Use the Session Info tab to analyze node-level issues such as bottlenecks or loops by observing spikes and trends in transaction counts.
- Traffic monitoring: Identify peak system usage times by selecting data points showing transactions within 10-minute intervals and monitoring logged-in user trends.
- Node performance comparison: Use the Group by Performance option to determine if performance degradation is isolated or widespread and compare metrics to the 1-Day Moving Average.
- Slow Patterns tab: Investigate anomalies in execution counts and high execution times when no obvious issues appear in other metrics.
Key Outcomes
- Proactively detect and diagnose performance bottlenecks and unusual transaction loads on your instance.
- Understand the correlation between user activity and system responsiveness to better manage instance health.
- Leverage detailed node-level analysis to pinpoint specific areas causing performance degradation.
- Transition smoothly from Application Insights to the recommended Instance Observer solution for ongoing monitoring.
Maintain the health and performance of your instance by monitoring key metrics related to users and transactions through the Application Insights Users and Transactions graph.
Starting with the Zurich release, Application Insights is no longer deployed, enhanced, or supported. It is recommended to evaluate the Overview of Instance Observer product available with the ServiceNow Impact packages. Work with your Account team to review Impact packages.
For details, see the Deprecation Process [KB0867184] article in the Now Support Knowledge Base.
You can detect anomalies and performance deviations in your instance by monitoring the relationship between users, transactions, and average transaction response time in the Users and Transactions graph on the Overview tab.
Use the Users and Transactions graph to do the following:
- Monitor the performance of transactions on your instance.
- View the transaction load on your instance.
- Track the number of logged-in users.
You access the graph by navigating to .
- Analyze system slowness by comparing the Transaction Count and Logged in Users figures. Note that a spike in the Transaction Count that has a corresponding spike in the Logged in Users count might be a false alarm.
- Determine whether a heavy transaction load is affecting system performance by comparing the Average Response Time figure to the Transaction Count amount. Look for spikes in the Transaction Count amount relative to the Average Response Time figure.
Dig deeper into a potential performance issue by drilling down to analyze issues at the node-level in the detail graphs on the Session Info tab.
- Look for spikes that indicate a bottleneck or loop condition in the Transaction Count detail graph. An upward trend line indicates the system is becoming overloaded. Select a data point on the graph to view a list of transactions in a 10-minute range.
- Look for spikes that indicate times of heaviest traffic on the instance in the Logged in Users detail graph. Monitor the trend lines to determine when the system is busiest.
- Determine whether performance degradation is happening on one node, across all nodes, or on a set of nodes that are consolidated using the Group by Performance option. Compare the node metrics to the 1-Day Moving Average amount. If you don’t see any obvious issues, look for anomalies in execution counts or high execution times on the Slow Patterns tab. If you see both, start your investigation there.