An experimental / beta version of the SDK is also available. It is built on top of OpenTelemetry, more information about beta version available here.

npm install applicationinsights@beta

Azure Application Insights monitors your backend services and components after you deploy them to help you discover and rapidly diagnose performance and other issues. Add this SDK to your Node.js services to include deep info about Node.js processes and their external dependencies such as database and cache services. You can use this SDK for your Node.js services hosted anywhere: your datacenter, Azure VMs and Web Apps, and even other public clouds.

This library tracks the following out-of-the-box:

• Incoming and outgoing HTTP requests
• Important system metrics such as CPU usage
• Unhandled exceptions
• Events from many popular third-party libraries (see Automatic third-party instrumentation)

You can manually track more aspects of your app and system using the API described in the Track custom telemetry section.

Important: On March 31st, 2025, support for instrumentation key ingestion will end. Instrumentation key ingestion will continue to work, but we’ll no longer provide updates or support for the feature. Transition to connection strings to take advantage of new capabilities.

1. Create an Application Insights resource in Azure by following these instructions.
2. Grab the Connection String from the resource you created in step 1. Later, you'll either add it to your app's environment variables or use it directly in your scripts.
3. Add the Application Insights Node.js SDK to your app's dependencies and package.json:

   npm install --save applicationinsights

   Note: If you're using TypeScript, please install @types/node package to prevent build issues, this npm package contains built-in typings.

4. As early as possible in your app's code, load the Application Insights package:

   let appInsights = require('applicationinsights');

5. Configure the local SDK by calling appInsights.setup('YOUR_CONNECTION_STRING');, using the connection string you grabbed in step 2. Or put it in the APPLICATIONINSIGHTS_CONNECTION_STRING environment variable and call appInsights.setup() without parameters.

   For more configuration options see below.

6. Finally, start automatically collecting and sending data by calling appInsights.start();.

Important: applicationinsights must be setup and started before you import anything else. There may be resulting telemetry loss if other libraries are imported first.

For out-of-the-box collection of HTTP requests, popular third-party library events, unhandled exceptions, and system metrics:

let appInsights = require("applicationinsights");
appInsights.setup("YOUR_CONNECTION_STRING").start();

• If the connection string is set in the environment variable APPLICATIONINSIGHTS_CONNECTION_STRING, .setup() can be called with no arguments. This makes it easy to use different connection strings for different environments.

Load the Application Insights library (i.e. require("applicationinsights")) as early as possible in your scripts, before loading other packages. This is needed so that the Application Insights library can prepare later packages for tracking. If you encounter conflicts with other libraries doing similar preparation, try loading the Application Insights library after those.

Because of the way JavaScript handles callbacks, additional work is necessary to track a request across external dependencies and later callbacks. By default this additional tracking is enabled; disable it by calling setAutoDependencyCorrelation(false) as described in the Configuration section below.

Auto correlation in Azure Functions is supported automatically starting in 2.4.0, if using previous version following code should be added to handle the correlation available in Azure Functions environment.

...

// Default export wrapped with Application Insights FaaS context propagation
export default async function contextPropagatingHttpTrigger(context, req) {
    // Start an AI Correlation Context using the provided Function context
    const correlationContext = appInsights.startOperation(context, req);

    // Wrap the Function runtime with correlationContext
    return appInsights.wrapWithCorrelationContext(async () => {
        const startTime = Date.now(); // Start trackRequest timer

        // Run the Function
        const result = await httpTrigger(context, req);

        // Track Request on completion
        appInsights.defaultClient.trackRequest({
            name: context.req.method + " " + context.req.url,
            resultCode: context.res.status,
            success: true,
            url: req.url,
            time: new Date(startTime),
            duration: Date.now() - startTime,
            id: correlationContext.operation.parentId,
        });
        appInsights.defaultClient.flush();

        return result;
    }, correlationContext)();
};

The appInsights object provides a number of methods to setup SDK behavior. They are listed in the following snippet with their default values.

let appInsights = require("applicationinsights");
appInsights.setup("<YOUR_CONNECTION_STRING>")
    .setAutoDependencyCorrelation(true)
    .setAutoCollectRequests(true)
    .setAutoCollectPerformance(true, true)
    .setAutoCollectExceptions(true)
    .setAutoCollectDependencies(true)
    .setAutoCollectConsole(true, false)
    .setUseDiskRetryCaching(true)
    .setAutoCollectPreAggregatedMetrics(true)
    .setSendLiveMetrics(false)
    .setAutoCollectHeartbeat(false)
    .setAutoCollectIncomingRequestAzureFunctions(true)
    .setInternalLogging(false, true)
    .setDistributedTracingMode(appInsights.DistributedTracingModes.AI_AND_W3C)
    .enableWebInstrumentation(false)
    .start();

Please review their descriptions in your IDE's built-in type hinting, or applicationinsights.ts for detailed information on what these control, and optional secondary arguments.

Note that by default setAutoCollectConsole is configured to exclude calls to console.log (and other console methods). By default, only calls to supported third-party loggers (e.g. winston, bunyan) will be collected. You can change this behavior to include calls to console methods by using setAutoCollectConsole(true, true).

Note that by default enableWebInstrumentation will use the connection string for SDK initialization. If you want to use a different one, you can set it as enableWebInstrumentation(true, "your-connection-string").

The TelemetryClient object contains a config property with many optional settings. These can be set as follows:

client.config.PROPERTYNAME = VALUE;

These properties are client specific, so you can configure appInsights.defaultClient separately from clients created with new appInsights.TelemetryClient().

All these properties except httpAgent, httpsAgent and aadTokenCredential could be configured using configuration file applicationinsights.json located under root folder of applicationinsights package installation folder, Ex: node_modules/applicationinsights. These configuration values will be applied to all TelemetryClients created in the SDK.

{
    "samplingPercentage": 80,
    "enableAutoCollectExternalLoggers": true,
    "enableAutoCollectExceptions": true,
    "enableAutoCollectHeartbeat": true,
    "enableSendLiveMetrics": true,
    ...
}
  

Custom JSON file could be provided using APPLICATIONINSIGHTS_CONFIGURATION_FILE environment variable.

process.env.APPLICATIONINSIGHTS_CONFIGURATION_FILE = "C:/applicationinsights/config/customConfig.json"

// Application Insights SDK setup....

Alternatively, instead of using a configuration file, you can specify the entire content of the JSON configuration via the environment variable APPLICATIONINSIGHTS_CONFIGURATION_CONTENT.

By default, the SDK will send all collected data to the Application Insights service. If you collect a lot of data, you might want to enable sampling to reduce the amount of data sent. Set the samplingPercentage field on the Config object of a Client to accomplish this. Setting samplingPercentage to 100 (the default) means all data will be sent, and 0 means nothing will be sent.

If you are using automatic correlation, all data associated with a single request will be included or excluded as a unit.

Add code such as the following to enable sampling:

const appInsights = require("applicationinsights");
appInsights.setup("<YOUR_CONNECTION_STRING>");
appInsights.defaultClient.config.samplingPercentage = 33; // 33% of all telemetry will be sent to Application Insights
appInsights.start();

If your application consists of multiple components that you wish to instrument all with the same Instrumentation Key and still see these components as separate units in the Portal as if they were using separate Instrumentation Keys (for example, as separate nodes on the Application Map) you may need to manually configure the RoleName field to distinguish one component's telemetry from other components sending data to your Application Insights resource. (See Monitor multi-component applications with Application Insights (preview))

Use the following to set the RoleName field:

const appInsights = require("applicationinsights");
appInsights.setup("<YOUR_CONNECTION_STRING>");
appInsights.defaultClient.context.tags[appInsights.defaultClient.context.keys.cloudRole] = "MyRoleName";
appInsights.start();

For node server with configuration enableWebInstrumentation set to true or environment variable APPLICATIONINSIGHTS_WEB_INSTRUMENTATION_ENABLED = true, web Instrumentation will be enabled on node server response when all of the following requirements are met:

• Response has status code 200.
• Response method is GET.
• Sever response has Content-Type html.
• Server response must have both <head> and </head> Tags.
• If response is compressed, it must have only one Content-Encoding type, and encoding type must be one of gzip, br or deflate.
• Response does not contain current /backup web Instrumentation CDN endpoints. (current and backup Web Instrumentation CDN endpoints here)

web Instrumentation CDN endpoint can be changed by setting environment variable APPLICATIONINSIGHTS_WEB_INSTRUMENTATION_SOURCE = "web Instrumentation CDN endpoints". web Instrumentation connection string can be changed by setting environment variable APPLICATIONINSIGHTS_WEB_INSTRUMENTATION_CONNECTION_STRING = "web Instrumentation connection string"

Note: web Instrumentation may slow down server response time, especially when response size is large or response is compressed. For the case in which some middle layers are applied, it may result in web Instrumentation not working and original response will be returned.

In order to track context across asynchronous calls, some changes are required in third party libraries such as mongodb and redis. By default ApplicationInsights will use diagnostic-channel-publishers to monkey-patch some of these libraries. This can be disabled by setting the APPLICATION_INSIGHTS_NO_DIAGNOSTIC_CHANNEL environment variable. Note that by setting that environment variable, events may no longer be correctly associated with the right operation. Individual monkey-patches can be disabled by setting the APPLICATION_INSIGHTS_NO_PATCH_MODULES environment variable to a comma separated list of packages to disable, e.g. APPLICATION_INSIGHTS_NO_PATCH_MODULES=console,redis to avoid patching the console and redis packages.

The following modules are available: azuresdk, bunyan, console, mongodb, mongodb-core, mysql, redis, winston, pg, and pg-pool. Visit the diagnostic-channel-publishers' README for information about exactly which versions of these packages are patched.

Automatic instrumentation for several Azure SDKs is also enabled, currently Cognitive Search, Communication Common and Cosmos DB SDKs are not supported. Javascript Azure SDKs

The bunyan, winston, and console patches will generate Application Insights Trace events based on whether setAutoCollectConsole is enabled. The rest will generate Application Insights Dependency events based on whether setAutoCollectDependencies is enabled. Make sure that applicationinsights is imported before any 3rd-party packages for them to be instrumented successfully.

To enable sending live metrics of your app to Azure, use setSendLiveMetrics(true). Filtering of live metrics in the Portal is currently not supported.

Note: The ability to send extended native metrics was added in version 1.4.0

To enable sending extended native metrics of your app to Azure, simply install the separate native metrics package. The SDK will automatically load it when it is installed and start collecting Node.js native metrics.

npm install applicationinsights-native-metrics

Currently, the native metrics package performs autocollection of Garbage Collection CPU time, Event Loop ticks, and heap usage:

• Garbage Collection: The amount of CPU time spent on each type of garbage collection, and how many occurrences of each type.
• Event Loop: How many ticks occurred and how much CPU time was spent in total.
• Heap vs Non-Heap: How much of your app's memory usage is in the heap or non-heap.

By default, this SDK will send headers understood by other applications/services instrumented with an Application Insights SDK. You can optionally enable sending/receiving of W3C Trace Context headers in addition to the existing AI headers, so you will not break correlation with any of your existing legacy services. Enabling W3C headers will allow your app to correlate with other services not instrumented with Application Insights, but do adopt this W3C standard.

const appInsights = require("applicationinsights");
appInsights
  .setup("<YOUR_CONNECTION_STRING>")
  .setDistributedTracingMode(appInsights.DistributedTracingModes.AI_AND_W3C)
  .start()

You can track any request, event, metric or exception using the Application Insights client. Examples follow:

let appInsights = require("applicationinsights");
appInsights.setup().start(); // assuming connection string is in environment variables. start() can be omitted to disable any non-custom data
let client = appInsights.defaultClient;
client.trackEvent({name: "my custom event", properties: {customProperty: "custom property value"}});
client.trackException({exception: new Error("handled exceptions can be logged with this method")});
client.trackMetric({name: "custom metric", value: 3});
client.trackTrace({message: "trace message"});
client.trackDependency({target:"http://dbname", name:"select customers proc", data:"SELECT * FROM Customers", duration:231, resultCode:0, success: true, dependencyTypeName: "ZSQL"});
client.trackRequest({name:"GET /customers", url:"http://myserver/customers", duration:309, resultCode:200, success:true});
 
let http = require("http");
http.createServer( (req, res) => {
  client.trackNodeHttpRequest({request: req, response: res}); // Place at the beginning of your request handler
});

Note that custom properties are converted to their string representation before being sent, see Using properties for more information.

An example utility using trackMetric to measure how long event loop scheduling takes:

function startMeasuringEventLoop() {
  var startTime = process.hrtime();
  var sampleSum = 0;
  var sampleCount = 0;

  // Measure event loop scheduling delay
  setInterval(() => {
    var elapsed = process.hrtime(startTime);
    startTime = process.hrtime();
    sampleSum += elapsed[0] * 1e9 + elapsed[1];
    sampleCount++;
  }, 0);

  // Report custom metric every second
  setInterval(() => {
    var samples = sampleSum;
    var count = sampleCount;
    sampleSum = 0;
    sampleCount = 0;

    if (count > 0) {
      var avgNs = samples / count;
      var avgMs = Math.round(avgNs / 1e6);
      client.trackMetric({name: "Event Loop Delay", value: avgMs});
    }
  }, 1000);
}

public addTelemetryProcessor(telemetryProcessor: (envelope: Contracts.Envelope, context: { http.RequestOptions, http.ClientRequest, http.ClientResponse, Error, correlationContext }) => boolean)

You can process and filter collected data before it is sent for retention using Telemetry Processors. Telemetry processors are called one by one in the order they were added before the telemetry item is sent to the cloud.

If a telemetry processor returns false that telemetry item will not be sent.

All telemetry processors receive the telemetry data and its envelope to inspect and modify. They also receive a context object. The contents of this object is defined by the contextObjects parameter when calling a track method for manually tracked telemetry. For automatically collected telemetry, this object is filled with available request information and the persistent request context as provided by appInsights.getCorrelationContext() (if automatic dependency correlation is enabled).

The TypeScript type for a telemetry processor is:

telemetryProcessor: (envelope: ContractsModule.Contracts.Envelope, context: { http.RequestOptions, http.ClientRequest, http.ClientResponse, Error, correlationContext }) => boolean;

For example, a processor that removes stack trace data from exceptions might be written and added as follows:

function removeStackTraces ( envelope, context ) {
  if (envelope.data.baseType === "ExceptionData") {
    var data = envelope.data.baseData;
    if (data.exceptions && data.exceptions.length > 0) {
      for (var i = 0; i < data.exceptions.length; i++) {
        var exception = data.exceptions[i];
        exception.parsedStack = null;
        exception.hasFullStack = false;
      }
    }
    // Add extra properties
    var originalError = context["Error"];
    if(originalError && originalError.prop){
      data.properties = data.properties || {};
      data.properties.customProperty = originalError.prop;
    }
  }
  return true;
}

appInsights.defaultClient.addTelemetryProcessor(removeStackTraces);

More info on the telemetry API is available in the docs.

You can create multiple Azure Application Insights resources and send different data to each by using their respective connection string. For example:

let appInsights = require("applicationinsights");

// configure auto-collection under one Connection String
appInsights.setup("<YOUR_CONNECTION_STRING>").start();

// track some events manually under another connection string
let otherClient = new appInsights.TelemetryClient("<YOUR_CONNECTION_STRING>");
otherClient.trackEvent({name: "my custom event"});

• Track dependencies

  let appInsights = require("applicationinsights");
  let client = new appInsights.TelemetryClient();
  
  var success = false;
  let startTime = Date.now();
  // execute dependency call here....
  let duration = Date.now() - startTime;
  success = true;
  
  client.trackDependency({target:"http://dbname", name:"select customers proc", data:"SELECT * FROM Customers", duration:duration, resultCode:0, success: true, dependencyTypeName: "ZSQL"});

• Assign custom properties to be included with all events

  appInsights.defaultClient.commonProperties = {
    environment: process.env.SOME_ENV_VARIABLE
  };

• Manually track all HTTP GET requests

  Note that all requests are tracked by default. To disable automatic collection, call .setAutoCollectRequests(false) before calling start().

  appInsights.defaultClient.trackRequest({name:"GET /customers", url:"http://myserver/customers", duration:309, resultCode:200, success:true});

  Alternatively you can track requests using trackNodeHttpRequest method:

  var server = http.createServer((req, res) => {
    if ( req.method === "GET" ) {
        appInsights.defaultClient.trackNodeHttpRequest({request:req, response:res});
    }
    // other work here....
    res.end();
  });

• Track server startup time

  let start = Date.now();
  server.on("listening", () => {
    let duration = Date.now() - start;
    appInsights.defaultClient.trackMetric({name: "server startup time", value: duration});
  });

"Self-diagnostics" refers to internal logging from Application Insights Node.js SDK.

This functionality can be helpful for spotting and diagnosing issues with Application Insights itself.

By default, Application Insights Node.js SDK logs at warning level to console, following code demonstrate how to enable debug logging as well and generate telemetry for internal logs:

let appInsights = require("applicationinsights");
appInsights.setup("<YOUR_CONNECTION_STRING>")
    .setInternalLogging(true, true) // Enable both debug and warning logging
    .setAutoCollectConsole(true, true) // Generate Trace telemetry for winston/bunyan and console logs
    .start();

Debug Logs could be enabled as well using APPLICATION_INSIGHTS_ENABLE_DEBUG_LOGS environment variable, and APPLICATION_INSIGHTS_DISABLE_WARNING_LOGS environment variable to disable warnings. Logs could be put into local file using APPLICATIONINSIGHTS_LOG_DESTINATION environment variable, supported values are file and file+console, a file named applicationinsights.log will be generated on tmp folder by default, including all logs, /tmp for *nix and USERDIR/AppData/Local/Temp for Windows. Log directory could be configured using APPLICATIONINSIGHTS_LOGDIR environment variable.

process.env.APPLICATIONINSIGHTS_LOG_DESTINATION = "file";
process.env.APPLICATIONINSIGHTS_LOGDIR = "C:/applicationinsights/logs"

// Application Insights SDK setup....

• Ongoing development takes place on the develop branch. Please submit pull requests to this branch.
• Releases are merged to the master branch and published to npm.

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repositories using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.

As this SDK is designed to enable applications to perform data collection which is sent to the Microsoft collection endpoints the following is required to identify our privacy statement.

The software may collect information about you and your use of the software and send it to Microsoft. Microsoft may use this information to provide services and improve our products and services. You may turn off the telemetry as described in the repository. There are also some features in the software that may enable you and Microsoft to collect data from users of your applications. If you use these features, you must comply with applicable law, including providing appropriate notices to users of your applications together with a copy of Microsoft’s privacy statement. Our privacy statement is located at https://go.microsoft.com/fwlink/?LinkID=824704. You can learn more about data collection and use in the help documentation and our privacy statement. Your use of the software operates as your consent to these practices.

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft’s Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies.

MIT