VOOZH about

URL: https://docs.datadoghq.com/profiler/

⇱ Continuous Profiler

For AI agents: A markdown version of this page is available at https://docs.datadoghq.com/profiler.md. A documentation index is available at /llms.txt.

Continuous Profiler


Find CPU, memory, and IO bottlenecks, broken down by method name, class name, and line number, to significantly reduce end-user latency and infrastructure costs.

Low impact in production

Continuous Profiler runs in production across all services by using technologies such as JDK Flight Recorder to have minimal impact on your host’s CPU and memory usage.

Getting started

Profiling your service to visualize all your stack traces in one place takes just minutes.

Instrument your application

Guide to using the profiler

The Getting Started with Profiler guide takes a sample service with a performance problem and shows you how to use Continuous Profiler to understand and fix the problem.

Explore Datadog profiler

After you configure your application to send profiles to Datadog, start getting insights into your code performance.

By default, profiles are retained for eight days, and metrics generated from profile data are retained for one month.

👁 learning center
Try Diagnose Code Performance Issues in the Learning Center

The Datadog Learning Center is full of hands-on courses to help you learn about this topic. Enroll at no cost to investigate and improve application code performance in production with Datadog Continuous Profiler.

ENROLL NOW

Profile types

See Profile Types for descriptions of the kinds of profile data collected for each supported language.

Search profiles by tags

Use tags to search profiles across any dimension—whether it’s a specific host, service, version, or any combination.

Track function performance over deployments

Obtain key profiling metrics from services such as top CPU usage by method, top memory allocations by thread, and CPU usage by version to visualize in your dashboards.

Connect traces to profiling data

Application processes that have both APM distributed tracing and continuous profiler enabled are automatically linked, so you can move directly from span information to profiling data on the Profiles tab to find specific lines of code related to performance issues.

Find changes in performance by comparing profiles

Comparing similar profiles from different times, environments, or deployments can help you understand the possible causes of and solutions to performance problems. The Datadog profiler offers comparison visualizations to make sense of why profiles are different based on time frames or tags that you scope by.

Further Reading

Additional helpful documentation, links, and articles:

Enable continuous profiler for your applicationDOCUMENTATION Getting Started with Continuous ProfilerDOCUMENTATION Learn more about available profile typesDOCUMENTATION Data collection and resolutionDOCUMENTATION Focus on code that matters with source code previews in Continuous ProfilerBLOG Introducing always-on production profiling in DatadogBLOG Datadog GitHub Action for continuous vulnerability analysisBLOG Compare and optimize your code with Datadog Profile Comparison.BLOG How we optimized our Akka application using Datadog's Continuous ProfilerBLOG Analyze Ruby code performance with Datadog Continuous ProfilerBLOG How our Cloud SIEM team uses context attributes with Continuous Profiler to get crucial performance insightsBLOG Making profiling visualizations accessible to engineers at all levelsBLOG Why continuous profiling is the fourth pillar of observabilityBLOG Monitor your Kubernetes operators to keep applications running smoothlyBLOG Troubleshoot faster with the GitLab Source Code integration in DatadogBLOG Optimize Request Latency with Profiling TimelineLEARNING CENTER