Use geoprocessing services in web apps

A geoprocessing service can be consumed in a custom web app, providing the clients of your web app with powerful geoprocessing and analysis solutions. Generally speaking, creating a web app can be considered a developer (or programmer) task. A web app can be built from scratch using the JavaScript API. Alternatively, a nondeveloper can deploy a configurable web app using Web AppBuilder in ArcGIS Enterprise.

Regardless of the method you use to build a web app, the communication of the underlying geoprocessing service occurs through REST and connects to the service end point on ArcGIS Server.

  • The service end point is a URL, typically in the following format: http://<server>/arcgis/rest/<ServiceName>
  • Each service end point has information that describes the service, the operations that can be performed, and the resources that compose the service. In the geoprocessing context, one or more tools compose the resources of a service.

To learn more about using web tools with ArcGIS REST API, see Geoprocessing services.

Use geoprocessing services with custom web apps

Typically, a custom web app is constructed using ArcGIS API for JavaScript. Although you can construct any app that can communicate with the service over REST, the following discussion focuses on JavaScript. If you're already comfortable with JavaScript and geoprocessing, you may want to review the samples and code snippets. Otherwise, the remainder of this section outlines the following process to add geoprocessing to your web app:

  1. Initializing the geoprocessing service
  2. Setting up input parameters
  3. Executing
  4. Getting and handling the result or results

The following JavaScript code is taken from Calculate Viewshed sample with only the relevant pieces to demonstrate the four steps required to execute geoprocessing tools. The sample in the JavaScript help provides a complete example, whereas the following only produces a working web page with additional code:

// esri.tasks.Geoprocessor is required for using Geoprocessor.
//  Add it along with other dojo.require statements.

// Step 1: Initialize the geoprocessing and point to the REST URL
var gpUrl = "";
var gp = new Geoprocessor(gpUrl);	

// Set output spatial reference 
gp.outSpatialReference = { wkid: 102100 };

function computeViewshed(event) { 

   // Step 2: Set up input parameters and assign or collect inputs from user
   var point = new Point({ 
        longitude: event.mapPoint.longitude,
        latitude: event.mapPoint.latitude  

   var inputGraphic = new Graphic({
      geometry: point, 
      symbol: markerSymbol 

   var inputGraphicContainer = [];
   var featureSet = new FeatureSet();
   featureSet.features = inputGraphicContainer;  

   var vsDistance = new LinearUnit();  
   vsDistance.distance = 5;
   vsDistance.units = "miles";  
   var params = {  
      "Input_Observation_Point": featureSet,
      "Viewshed_Distance": vsDistance  

   // Step 3: Execute the service. In this instance, the service is Synchronous; Execute
   // will be called. An Asynchronous service will use gp.submitJob.

// Step 4: Render the result. Symbology is assigned and the graphics are added to the map.
function drawResultData(result) { 
   var resultFeatures = result.results[0].value.features;  
   // Assign each resulting graphic a symbol  
   var viewshedGraphics = { 
      feature.symbol = fillSymbol;     
      return feature; 
   // Add the resulting graphics to the graphics layer  

In the sample above, the Geoprocessor class must be loaded, and the REST URL to the service is used with a global variable to define the geoprocessing task. The next step is to construct the parameters specific to the tool the app will be running. This example requires a point (feature) input and a distance (linear unit). The REST URL specific to the tool in the service will define input parameters. A service will be set up as either synchronous or asynchronous, and this setting controls whether the execution is done as submit GP job or execute GP task. The output for each tool is also explained on the same page as the input parameters; information about the type and how it can be displayed is given.

For a detailed explanation of these steps, see Use geoprocessing tasks in the ArcGIS API for JavaScript.


The previous example returns features representing the locations an individual would be able to see on the earth from the viewshed analysis. A geoprocessing service can provide text, numbers, or file output, that is, items that cannot be drawn on a map. In these scenarios, you need to take a different approach, and display the output from the service in a manner that is appropriate to the app you are building. For example, a pop-up box with text output or a link to save a file may suffice.

Use geoprocessing services with Web AppBuilder

To use Web AppBuilder, you do not need to write code; you only need to know the service endpoint (REST URL) of your geoprocessing service. A geoprocessing service is used in a web app created by Web AppBuilder by configuring the geoprocessing widget. You first need to create a web app in your portal. Once you have selected the style and map, you can add a new geoprocessing widget and configure it. The widget configuration allows you to set input and output options as they relate to the specific service. These input and output parameters that define the service dictate which options are available through the widget configuration. For example, the widget configuration allows you to control the type of feature (square, circle, freehand, and so on) that a user can enter if the geoprocessing service was published to accept feature input. The same premise applies to the output. A geoprocessing service set up to output features allows you to configure the symbology and how the web app should render the geoprocessing service output. A service that has been configured with a result map layer will have the result added to your web app and be symbolized based on settings chosen during the initial publishing process.