Node.js Performance APIs: An Introduction

Damilola Olatunji

Damilola Olatunji on

Node.js Performance APIs: An Introduction

Node.js offers a rich set of performance data through its APIs, similar to how modern web browsers expose performance metrics for webpages.

With tools like the Performance Measurement API and the perf_hooks module in Node.js, you can gain insights into your application's performance from the server's perspective, closely aligning with what the end user experiences.

In this article, the first of a two-part series about performance hooks, we'll explore how to approach performance optimization in Node.js. Along the way, we'll leverage Node's built-in timing APIs to enhance the efficiency of our application.

Let's get going!


To follow along with this tutorial, you'll need a basic knowledge of Node.js and a recent version installed, preferably the latest LTS.

Understanding Fundamental Node.js Performance API Concepts

Before diving into Node.js performance APIs, it's important to understand some essential terms that frequently appear in discussions and documentation: a monotonic clock, high resolution time API, and performance timeline. Let's look at each of these in turn.

1. A Monotonic Clock

In JavaScript, the function generates timestamps based on the system clock, counting from the Unix epoch (1970-01-01T00:00:00Z). However, using these timestamps for precise measurements can be unreliable because they are vulnerable to any alterations in the system clock.

In contrast, a monotonic clock offers a timekeeping mechanism that always progresses forward, without any backward movement. Its key advantage lies in its stability against system clock changes, making it ideal for accurately measuring the time span between events or the duration of specific tasks.

2. High Resolution Time API

The High Resolution Time API provides a JavaScript interface for obtaining highly accurate time measurements. It uses a sub-millisecond monotonic timestamp that remains unaffected by any changes to the system clock.

This API's functionality is encapsulated in the method. It returns a timestamp in sub-milliseconds, calculated relative to performance.timeOrigin — a UNIX timestamp denoting the initiation of a current Node.js process.

Here's a quick example to illustrate its usage:

import { performance } from "node:perf_hooks"; console.log(;

Running this code produces output such as:


This means that the console.log statement runs about 26 milliseconds from the initialization of the Node.js process.

Whenever you come across terms like "high resolution" or "high precision timestamp" in JavaScript, this refers to values generated by the method.

3. Performance Timeline

The Performance Timeline in Node.js is a structured sequence of timing details and key performance milestones. These are captured both from the Node.js process itself and any code executed within your application.

This timeline consists of PerformanceEntry objects (and subclasses), each offering insights into the type and name of an event, along with precise timestamps indicating both the start and the duration of the event.

Here's an example of such entries:

[ PerformanceMark { name: 'query_start', entryType: 'mark', startTime: 49.083899974823, duration: 0, detail: null }, PerformanceMark { name: 'query_end', entryType: 'mark', startTime: 674.0229699611664, duration: 0, detail: null } ]

The Performance Timeline is mainly used to monitor and extract comprehensive performance metrics throughout the entire lifecycle of a Node.js application.

This information is invaluable for real-time performance analysis and can be seamlessly integrated with monitoring systems like AppSignal for Node.js to enhance application performance insights and diagnostics.

Inside the Node.js Performance API

Node.js offers a suite of Performance Measurement APIs within its perf_hooks module that are specifically designed to record Node.js performance data.

These APIs are inspired by, and partially implement, the Web Performance APIs found in browser environments. However, they are tailored and adapted to suit server-side programming contexts.

In the following sections, we'll dive into some of the key APIs provided by this module:

  1. PerformanceNodeTiming
  3. Performance.mark()
  4. Performance.getEntries()
  5. Performance.measure()
  6. Performance.clearMarks()
  7. Performance.clearMeasures()
  8. PerformanceResourceTiming

1. PerformanceNodeTiming

In any Node.js program, the initial entry in the Performance Timeline is a PerformanceNodeTiming instance. It distinctively captures the timestamps of significant milestones during the startup phase of a process. To access this data, you can print the performance object, as demonstrated below:

import { performance } from "node:perf_hooks"; console.log(performance);

This results in an output like:

Performance { nodeTiming: PerformanceNodeTiming { name: 'node', entryType: 'node', startTime: 0, duration: 17.71333795785904, nodeStart: 0.256866991519928, v8Start: 2.246299982070923, bootstrapComplete: 9.73558497428894, environment: 6.018610000610352, loopStart: 13.11082899570465, loopExit: -1, idleTime: 0.03074 }, timeOrigin: 1700776448830.537 }

The nodeTiming object, as part of the Node.js runtime timings, provides a detailed view of various stages in the process lifecycle. Here's a breakdown of its key properties:

  • name: Identifies the performance entry, which is the Node.js process in this case.
  • entryType: Specifies the type of performance entry.
  • startTime: Indicates when the measurement started, with 0 denoting the beginning of the process execution.
  • duration: Represents the duration from the start of the process to the point when the entry was recorded.
  • nodeStart: A high resolution timestamp marking the start of the Node.js runtime initialization.
  • v8Start: The time at which the V8 JavaScript engine began initializing.
  • bootstrapComplete: The time taken for the Node.js environment to complete setup and get ready to execute user scripts.
  • environment: The duration required to establish the Node.js environment.
  • loopStart: The high resolution timestamp marking the start of the event loop.
  • loopExit: Indicates when the event loop exited; a value of -1 suggests it was still active at the time of recording.
  • idleTime: The amount of time the event loop spent idle, not engaged in processing tasks.


To measure an operation's execution time in Node.js, you can use to generate high-resolution timestamps.

By capturing two timestamps — one before the operation starts and another after its completion — and then calculating the difference between them, you can determine an operation's duration.

The following example demonstrates this:

import { performance } from "node:perf_hooks"; const t0 =; await fetch(""); const t1 =; console.log(`Fetch request took ${t1 - t0} milliseconds.`);

Running this script typically results in an output like:

Fetch request took 682.8348079919815 milliseconds.

While this works for simple cases, there are more flexible methods available, which we'll explore later in this tutorial.

3. Performance.mark()

The performance.mark() function is a powerful method for creating and adding named markers to the Performance Timeline. These markers can later be retrieved for performance analysis.

For instance, to track the start and end of an HTTP request, you might use it as follows:

import { performance } from "node:perf_hooks"; const mark1 = performance.mark("mark_fetch_start"); await fetch(""); const mark2 = performance.mark("mark_fetch_end"); console.log(mark1); console.log(mark2);

This will result in an output similar to:

PerformanceMark { name: 'mark_fetch_start', entryType: 'mark', startTime: 17.95012092590332, duration: 0, detail: null } PerformanceMark { name: 'mark_fetch_end', entryType: 'mark', startTime: 959.4822289943695, duration: 0, detail: null }

In this example, the startTime for each PerformanceMark object reflects a high-resolution timestamp, akin to what provides. The duration for such objects is always zero since they represent specific moments in time.

You can also enrich these markers with additional information using the detail property:

import { performance } from "node:perf_hooks"; const url = ""; const options = { detail: { url, }, }; const mark1 = performance.mark("mark_fetch_start", options); await fetch(url); const mark2 = performance.mark("mark_fetch_end", options); console.log(mark1); console.log(mark2);

Resulting in:

// outputs: PerformanceMark { name: 'mark_fetch_start', entryType: 'mark', startTime: 25.053369998931885, duration: 0, detail: { url: '' } } PerformanceMark { name: 'mark_fetch_end', entryType: 'mark', startTime: 690.2271919250488, duration: 0, detail: { url: '' } }

By adding the detail property, the performance marks provide not just timing information but also contextual data, enhancing the comprehensiveness of your performance analysis.

4. Performance.getEntries()

A PerformanceEntry is a unique entity representing a single performance metric within an application's Performance Timeline. For instance, the PerformanceMark objects we discussed earlier are subclasses of PerformanceEntry, characterized by an entryType of mark.

Use getEntries() to retrieve an array of all the PerformanceEntry objects logged up to a specific moment in your program's execution. Here's an example:

import { performance } from "node:perf_hooks"; performance.mark("mark_fetch_start"); await fetch(""); performance.mark("mark_fetch_end"); console.log(performance.getEntries());

Running this snippet of code produces an array output like this:

[ PerformanceMark { name: 'mark_fetch_start', entryType: 'mark', startTime: 49.083899974823, duration: 0, detail: null }, PerformanceMark { name: 'mark_fetch_end', entryType: 'mark', startTime: 674.0229699611664, duration: 0, detail: null } ]

For situations where your Performance Timeline contains numerous entries and you wish to access only specific subsets, Node.js offers getEntriesByName() and getEntriesByType() methods. These methods allow for filtering based on the name and entryType properties of the performance entries:

// get all entries whose name is `mark_fetch_start` console.log(performance.getEntriesByName("mark_fetch_start")); // get all entries whose name is `mark_fetch_start` and whose type is `mark` console.log(performance.getEntriesByName("mark_fetch_start", "mark")); // get all entries whose entryType is `mark` console.log(performance.getEntriesByType("mark"));

These methods provide a more focused approach to examining performance data, enabling you to isolate and analyze specific aspects of your application's performance.

5. Performance.measure()

The Performance.measure() method creates PerformanceMeasure objects that quantify the duration between two specific points in time (often marked by PerformanceMark objects) within an application's Performance Timeline:

performance.measure(name[, startMarkOrOptions[, endMark]])

If only the name parameter is supplied, it will generate a PerformanceMeasure object capturing the program's duration up until the call:


This command results in an output like:

PerformanceMeasure { name: 'measureProgram', entryType: 'measure', startTime: 0, duration: 936.0637300014496 }

A more practical usage of measure() is calculating the elapsed time between two previously recorded PerformanceMark objects:

import { performance } from "node:perf_hooks"; const url = ""; const mark1 = performance.mark("mark_fetch_start"); await fetch(url); const mark2 = performance.mark("mark_fetch_end"); console.log( performance.measure("measureProgram", "mark_fetch_start", "mark_fetch_end") );

This will output something similar to the following:

PerformanceMeasure { name: 'measureFetch', entryType: 'measure', startTime: 22.075212001800537, duration: 903.0410770177841 }

In this scenario, startTime corresponds to the start time of the first mark (mark_fetch_start), and duration is the difference in time between the two marks.

This effectively measures the time that the HTTP request takes, offering a precise and practical way to assess specific durations within your application's performance timeline.

6. Performance.clearMarks()

The performance.clearMarks() function removes PerformanceMark objects from the Performance Timeline. This feature provides flexibility, allowing you to either clear all existing marks or remove specific ones identified by a name.

This is particularly useful for managing and resetting performance data, especially in long-running applications or during repeated testing phases.

The function can be employed as follows:

// Remove all `PerformanceMark` objects performance.clearMarks(); // Remove only `PerformanceMark` objects named `mark_fetch_start` performance.clearMarks("mark_fetch_start");

7. Performance.clearMeasures()

This function does the same thing as clearMarks() but removes PerformanceMeasure objects instead.

8. PerformanceResourceTiming

The PerformanceResourceTiming API is a specialized tool for capturing detailed network timing data, allowing you to track the duration of network requests in your code.

While we've used methods like, performance.mark(), and performance.measure() to track the duration of an HTTP request, they are more general-purpose tools.

In practice, the fetch API automatically gathers performance timings for each request, which are then recorded in the Performance Timeline as PerformanceResourceTiming objects.

import { performance } from "node:perf_hooks"; const url = ""; const response = await fetch(url); await response.json(); console.log(performance.getEntries());

Executing this script will result in an output resembling:

[ PerformanceResourceTiming { name: '', entryType: 'resource', startTime: 33.411154985427856, duration: 1665.9909150600433, initiatorType: 'fetch', nextHopProtocol: undefined, workerStart: 0, redirectStart: 0, redirectEnd: 0, fetchStart: 33.411154985427856, domainLookupStart: undefined, domainLookupEnd: undefined, connectStart: undefined, connectEnd: undefined, secureConnectionStart: undefined, requestStart: 0, responseStart: 0, responseEnd: 1699.4020700454712, transferSize: 300, encodedBodySize: 0, decodedBodySize: 0 } ]

This PerformanceResourceTiming object encapsulates several measurements for the HTTP request. Key attributes include:

  • startTime, fetchStart: The high resolution timestamp marking the commencement of the resource fetch.
  • responseEnd: The high resolution timestamp right after Node.js receives the last byte of the resource (or just before the transport connection closes, whichever occurs first).
  • duration: The difference between responseEnd and startTime.
  • transferSize: The total size of the fetched resource, including both response headers and body, measured in octets (commonly the same as bytes).

It's important to note that this data gets recorded in the timeline only when the response is actually used. Omitting the await response.json() line, for instance, would result in an empty array.

Finally, let's take a quick look at how to manage resource buffer sizes.

Managing Resource Buffer Sizes in Node

The Performance Timeline has a default capacity limit for PerformanceResourceTiming objects, set at 250. To adjust this limit, you can utilize the setResourceTimingBufferSize() method:


This command increases the buffer size, allowing up to 500 PerformanceResourceTiming objects to be stored simultaneously in the Performance Timeline.

If you need to clear all the existing PerformanceResourceTiming objects from the timeline, the clearResourceTimings() method comes in handy:


You can also listen for the resourcetimingbufferfull event to take actions when the performance resource timing buffer is full. For example, you could increase the buffer capacity or clear the buffer so that new entries can be added:

performance.addEventListener("resourcetimingbufferfull", (event) => { if (performance.getEntriesByType("resource").length < 1000) { performance.setResourceTimingBufferSize(1000); } else { performance.clearResourceTimings(); } });

And that's it for now!

Up Next: Practical Uses of Node.js Performance APIs

In this blog post, we focused on examining the most important Node.js Performance APIs and their various uses.

In the next and final part of this series, we'll dive into practical implementations of these APIs. I'll demonstrate their real-world value to optimize and resolve performance issues in your Node.js applications.

Stay tuned, and happy coding!

P.S. If you liked this post, subscribe to our JavaScript Sorcery list for a monthly deep dive into more magical JavaScript tips and tricks.

P.P.S. If you need an APM for your Node.js app, go and check out the AppSignal APM for Node.js.

Damilola Olatunji

Damilola Olatunji

Damilola is a freelance technical writer and software developer based in Lagos, Nigeria. He specializes in JavaScript and Node.js, and aims to deliver concise and practical articles for developers. When not writing or coding, he enjoys reading, playing games, and traveling.

All articles by Damilola Olatunji

Become our next author!

Find out more

AppSignal monitors your apps

AppSignal provides insights for Ruby, Rails, Elixir, Phoenix, Node.js, Express and many other frameworks and libraries. We are located in beautiful Amsterdam. We love stroopwafels. If you do too, let us know. We might send you some!

Discover AppSignal
AppSignal monitors your apps