Continuous Profiler | Datadog

Continuous Profiler

Optimize code performance in production

Datadog Continuous Profiler allows you to analyze code performance in production, across your entire stack with minimal overhead. Leverage code profiling to quickly detect and optimize the most resource-consuming methods or classes in your application, increasing code efficiency and reducing cloud provider costs.


Discover bottlenecks in your code at a glance

Continuously profile your code without affecting your production environment. Use tags and facets to filter call stacks by various attributes, such as method, thread, and package. Visualize different profiles—including CPU, memory allocation, lock, and I/O—and break them down by method name, class name, and line number to troubleshoot bottlenecks and significantly reduce end-user latency and infrastructure costs.


    Correlate code profiles and distributed traces

    Our unified platform enables seamless, two-way correlation between profiles and distributed traces with a click of a button. Within a profile, you can identify the most resource-intensive requests. Similarly, when investigating a slow request, you can pivot to the related profiles and find resource bottlenecks.


      Derive actionable insights from code analysis

      Leverage an automatic heuristic analysis and a summary of the main problem areas of your code to identify and address performance issues such as blocked threads, inefficient garbage collection, and memory leaks. For even more granular insight, you can view the complete code profiling analysis, broken down by categories such as code cache, class loading, and heap size.


        View profiler metrics to identify trends that can help you isolate underperforming hosts. By correlating metrics from your services, such as top methods by CPU, top methods by compilation time, and top methods by lock wait time, you can get a more comprehensive view of your application’s performance—and if you find any interesting trends, you can add any of these graphs to your custom dashboards or create machine-learning-based alerts for instant anomaly or outlier detection.