In APM, a transaction trace gives a detailed snapshot of a single transaction in your application. A transaction trace records the available function calls, database calls, and external calls. You can use transaction traces to troubleshoot performance issues and to get detailed low-level insight into how your app is working.
Here are the default rules that govern which transactions a New Relic agent traces:
- Over the minute-long harvest cycle, all transactions that violate the threshold (either four times your Apdex T value or a specific number of seconds) are added to a pool of transactions.
- At the end of that minute, the New Relic agent selects the slowest transaction in that pool and performs a transaction trace on it.
These are the general rules, but there are some agent-specific differences. For example:
- The Java agent collects occasional non-slow transactions.
- If one transaction is frequently traced, some agents will select other transactions to give a more diverse sampling.
If you don't see traces in your account, it's possible that no transactions meet the necessary criteria. In this situation, you can adjust transaction trace settings to ensure some transactions will be traced.
To configure or edit trace settings, see the procedures for:
To view transaction traces:
- Do one of the following:
- In the Transaction traces section, click transaction traces to view additional details.
You can bring your logs and application's data together to make troubleshooting easier and faster. With logs in context, you can see log messages related to your errors and traces directly in your app's UI.
- From the Transactions page, click on a trace to go to the Trace details page.
- From the trace details page, click See logs.
- To view details related to an individual log message, click directly on the message.
You can also see logs in context of your infrastructure data, such as Kubernetes clusters. No need to switch to another UI page.
If you expect to see transaction traces but do not see them, follow the troubleshooting procedures.
For more information about using transaction traces, slow queries, and service maps to solve performance issues, see Analyze performance issues.