Profile Types

Once profiling is enabled, the following profile types are collected for supported Java versions:

CPU

The time each method spent running on the CPU. It includes your code that runs in the JVM (for example, Java, Kotlin), but not JVM operations or native code called from within the JVM.

Allocations

The number of heap allocations made by each method, including allocations which were subsequently freed.
Requires: Java 11

Allocated Memory

The amount of heap memory allocated by each method, including allocations which were subsequently freed.
Requires: Java 11

Live Heap (v1.61.0+)

The objects and memory allocated by each method that have not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks. The profiler automatically uses the most accurate engine available for your JVM version.
Requires: Java 11+

Wall Time in Native Code

The elapsed time spent by each method. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the method is running.

Class Load

The number of classes loaded by each method.

Thrown Exceptions

The number of errors and exceptions thrown by each method, as well as their type.

File I/O

The time each method spent reading from and writing to files.

Lock

The time each method spent waiting for a lock.

Socket I/O

The time each method spent reading from and writing to socket I/O.

Once profiling is enabled, the following profile types are collected, depending on your Python version as noted:

Wall Time

The elapsed time used by each function. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the function is running.
Requires: Python 2.7+

Lock Wait Time

The time each function spent waiting for a lock.
Requires: Python 2.7+

Locked Time

The time each function spent holding a lock.
Requires: Python 2.7+

Lock Acquires

The number of times each function acquired a lock.
Requires: Python 2.7+

Lock Releases

The number of times each function released a lock.
Requires: Python 2.7+

CPU

The time each function spent running on the CPU, including Python and native code.
Requires: Python 2.7+, POSIX platform

Heap Live Size

The amount of heap memory allocated by each function that has not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.
Requires: Python 3.5+

Allocated Memory

The amount of heap memory allocated by each function, including allocations which were subsequently freed.
Requires: Python 3.5+

Allocations

The number of heap allocations made by each function, including allocations which were subsequently freed.
Requires: Python 3.5+

Once profiling is enabled, the following profile types are collected for supported Go versions:

CPU Time

The time each function spent running on the CPU. Off-CPU time such as waiting for Networking, Channels, Mutexes, and Sleep are not captured in this profile. See Mutex and Block profiles.

Allocations

The number of objects allocated by each function in heap memory during the profiling period (default: 60s), including allocations which were subsequently freed. Go calls this alloc_objects. Stack allocations are not tracked. This is useful for investigating garbage collection load. See also the note about how this measure changes in version 1.33.0 in Delta profiles.

Allocated Memory

The amount of heap memory allocated by each function during the profiling period (default: 60s), including allocations which were subsequently freed. Go calls this alloc_space. Stack allocations are not tracked. This is useful for investigating garbage collection load. See also the note about how this measure changes in version 1.33.0 in Delta profiles.

Heap Live Objects

The number of objects allocated by each function in heap memory that remain in use after garbage collection. Go calls this inuse_objects. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.

Heap Live Size

The amount of heap memory allocated by each function that remains in use after garbage collection. Under default settings (GOGC=100), this will typically represent ~50% of the RSS usage of the process. Go calls this inuse_space. Use this metric to review memory consumption and diagnose leaks. For more details about how Go manages memory, see Go memory metrics demystified and A Guide to the Go Garbage Collector.

Mutex

The time functions have been waiting on mutexes during the profiling period (default: 60s). The stack traces in this profile point the Unlock() operation that allowed another goroutine blocked on the mutex to proceed. Short mutex contentions using spinlocks are not captured by this profile, but can be seen in the CPU profile. See also the note about how this measure changes in version 1.33.0 in Delta profiles.

Block

The time functions have been waiting on mutexes and channel operations during the profiling period (default: 60s). Sleep, GC, Network, and Syscall operations are not captured by this profile. Blocking operations are only captured after they become unblocked, so this profile cannot be used to debug applications that appear to be stuck. For mutex contentions, the stack traces in this profile point to blocked Lock() operations. This tells you where your program is getting blocked, while the mutex profile tells you what part of your program is causing the contention. See Datadog’s Block Profiling in Go research for more in-depth information. See also the note about how this measure changes in version 1.33.0 in Delta profiles. Note: The block profiler can cause noticeable overhead for production workloads. If enabling it in production, prefer high rates (such as 100000000, which is 100 milliseconds) and look for signs of increased latency or CPU utilization.

Goroutines

A snapshot of the number of goroutines currently executing the same functions (both on-CPU and waiting off-CPU). An increasing number of goroutines between snapshots can indicate that the program is leaking goroutines. In most healthy applications this profile is dominated by worker pools and the number of goroutines they use. Applications that are extremely latency-sensitive and use a large number of goroutines (> 10,000) should be aware that enabling this profile requires stop-the-world pauses. The pauses occur only once every profiling period (default 60s) and normally last for around 1µsec per goroutine. Typical applications with a p99 latency SLO of around 100ms can generally ignore this warning. See Datadog’s Goroutine Profiling in Go research for more in-depth information.

Delta profiles

Once profiling is enabled, the following profile types are collected for supported Ruby versions:

CPU

The time each function spent running on the CPU, including Ruby and native code.

Wall Time

The elapsed time used by each function. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the function is running.

Allocations (v2.3.0+)

The number of objects allocated by each method during the profiling period (default: 60s), including allocations which were subsequently freed. This is useful for investigating garbage collection load.
Requires: Manual enablement

Heap Live Objects (Preview, v2.18.0+)

The number of objects allocated by each method in heap memory that have not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.
Requires: Ruby 3.1+ and manual enablement

Heap Live Size (Preview, v2.18.0+)

The amount of heap memory allocated by each method that has not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.
Requires: Ruby 3.1+ and manual enablement

GVL profiling (in Timeline) (v2.11.0+)

Records time when threads are prevented from working by other “noisy neighbor” threads, including background threads. This is useful for investigating latency spikes in the application when using the timeline visualization.
Requires: Ruby 3.2+

Once profiling is enabled, the following profile types are collected for supported Node.js versions:

CPU

The time each function spent running on the CPU, including JavaScript and native code.

CPU profiling is available on Linux and macOS. The feature is not available on Windows.

Wall Time

The elapsed time used by each function. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the function is running.

Heap Live Objects

The number of objects allocated by each function in heap memory that have not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.

Heap Live Size

The amount of heap memory allocated by each function that has not yet been garbage collected. This is useful for investigating the overall memory usage of your service and identifying potential memory leaks.

Deep stack traces in Heap Live Size profiles are truncated to 64 frames.

Allocated Memory (Preview)

The amount of heap memory allocated by each function, including allocations which were subsequently freed.
Requires: Node.js 26+ and DD_PROFILING_ALLOCATION_ENABLED=true

Allocations (Preview)

The number of heap allocations made by each function, including allocations which were subsequently freed.
Requires: Node.js 26+ and DD_PROFILING_ALLOCATION_ENABLED=true

Once profiling is enabled, the following profile types are collected for supported .NET versions:

Wall Time

The elapsed time spent in managed methods. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the method is running.

CPU (v2.15+)

The time each method spent running on the CPU.

Thrown Exceptions (v2.31+)

The number of caught or uncaught exceptions raised by each method, as well as their type and message.

Allocations (v3.28+)

The number and size of allocated objects by each method, as well as their type. For .NET Framework, the size is not available.
_Requires: .NET Framework (with Datadog Agent 7.51+ and v3.2+) / .NET 6+, but Datadog recommends .NET 10+ for more accurate sampling.

Lock (v2.49+)

The number of times threads are waiting for a lock and for how long.
Requires: .NET Framework (requires Datadog Agent 7.51+) / .NET 5+

Live Heap (v3.28+)

A subset of the allocated objects (with their class name) that are still in memory.
_Requires: .NET 7+ but Datadog recommends .NET 10+ for more accurate sampling.

Outgoing HTTP requests (in Timeline) (Preview, v3.19+)

Start and end of outgoing HTTP requests with the duration of the different phases (DNS, security handshake, socket, request/response) and possible unexpected redirections.
Requires: .NET 7+

Thread lifetime (in Timeline) (v3.19+)

Start and end of threads life to easily detect ThreadPool starvation and short lived threads.
Requires: .NET Framework (with Datadog Agent 7.51+ and v3.2+) / .NET 5+

Garbage Collector CPU consumption (v3.19+)

The time garbage collector’s threads spent running on the CPU.
Requires: .NET Framework (with Datadog Agent 7.51+ and v3.2+) / .NET 5+

Note: Before .NET 10, Allocations and Live Heap profiling might show bigger objects more than smaller ones due to the sampling algorithm used by the .NET runtime. Datadog recommends using .NET 10+ for more statistically correct results.

Once profiling is enabled, the following profile types are collected for supported PHP versions:

Wall Time

The elapsed time used by each function. Elapsed time includes time when code is running on CPU, waiting for I/O, and anything else that happens while the function is running.

CPU

Shows the time each function spent running on the CPU.

Allocations (v0.88+)

The number of allocations by each function during the profiling period (default: 67s), including allocations which were subsequently freed. Stack allocations are not tracked.
Note: Not available when JIT is active on PHP 8.0.0-8.1.20 and 8.2.0-8.2.7

Allocated memory (v0.88+)

The amount of heap memory allocated by each function during the profiling period (default: 67s), including allocations which were subsequently freed. Stack allocations are not tracked.
Note: Not available when JIT is active on PHP 8.0.0-8.1.20 and 8.2.0-8.2.7

Thrown Exceptions (v0.92+)

The number of caught or uncaught exceptions raised by each method, as well as their type.

File I/O (Preview, v1.7.2+)

The time each method spent reading from and writing to files, as well as the amount of bytes read from and written to files.

Socket I/O (Preview, v1.7.2+)

The time each method spent reading from and writing to a socket, as well as the amount of bytes read from and written to sockets.

Once profiling is enabled, the following profile types are collected for supported languages and versions:

CPU

The time each function spent running on the CPU.

Allocations

The number of allocations by each function during the profiling period (default: 59s), including allocations which were subsequently freed. Stack allocations are not tracked.

Allocated memory

The amount of heap memory allocated by each function during the profiling period (default: 59s), including allocations which were subsequently freed. Stack allocations are not tracked.

Once profiling is enabled, the following profile types are collected for supported languages and versions:

CPU Time (eBPF)

Time each method or function spent running on the CPU. In multi-threaded programs, CPU time can be greater than elapsed time: if 2 threads are running during 45s each, you’d see “eBPF CPU Time, 1m 30s per minute”.