@dynatrace/cordova-plugin
Advanced tools
This plugin gives you the ability to use the Dynatrace instrumentation in your hybrid application (Cordova, Ionic, ..). It uses the Mobile Agent, the JavaScript Agent. The Mobile Agent will give you all device specific values containing lifecycle informat
This plugin gives you the ability to use the Dynatrace instrumentation in your hybrid application (Cordova, Ionic, Capacitor ..). It uses the Mobile Agent and the JavaScript Agent. The Mobile Agent will give you all device specific values containing lifecycle information and the JavaScript Agent will allow you to manually instrument your JavaScript/TypeScript code out of the box (TypeScript definitions included). The JavaScript Agent will cover the network calls (depending on your used libraries) and will automatically detect them.
The dynatrace-cordova-plugin is the old version of this plugin. Consider it deprecated. Only use @dyntrace/cordova-plugin from now on, if you want to have the newest version. If you are upgrading from the old version have a look at our migration guide which is explaining what has changed. The versioning changed as well, the old 7.2.x was based on the versions of the Mobile Agents used. Now the plugin has its own versioning. The version of the used Mobile Agent can be seen here.
This agent versions are configured in this plugin:
To install the plugin in your Cordova based project you must enter the following command in the root directory of your cordova based project. E.g. :
cordova plugin add @dynatrace/cordova-plugin --save
If you want to instrument your Cordova application just go to your Dynatrace WebUI and select the menu point "Deploy Dynatrace". Choose to setup mobile monitoring and select Cordova. Afterwards it is possible for you to add the Web part (JavaScript Agent) automatically and download the dynatrace.config.js file.
This file should be placed in the root of your project (same place where the package.json is stored). If the file is not available the instrumentation will not work.
Ionic Webview for Cordova: If you are using the Ionic WebView for Cordova (cordova-plugin-ionic-webview) you need to make sure you set correct cookie domain.
After starting the Cordova or Ionic build, with ionic cordova build android or cordova build android the instrumentation will be handled by the plugin. Of course android can be substituted with ios platform. If you have trouble with not finding the .gradle, .plist or dynatrace.config.js file you can specify those via custom arguments. E.g. cordova build android --config=C:\config\dynatrace.config.js is a valid command and will try to use the configuration stored in the \config folder.
This section should explain what data should be visible after applying the plugin. The most important aspect is that there should be a combined session. We talk about a combined session when there are Mobile and Web actions within one session.
This screen shows what a combined sample user session should roughly look like. The session contains a Mobile action and some Web(Load, XHR) actions. Those first two actions should always be visible in your application session. The first one defines the native application startup ("Loading EasyTravel"). The second one defines the first load of the index.html of your hybrid application ("Loading of page ..").
To instrument your Ionic Capacitor app, you will need to follow the steps below:
npm install @dynatrace/cordova-pluginhybridWebView {
enabled true
domains 'http://localhost', '.yourotherdomains.com'
}
<platform> with ios or android):For Capacitor version < 3.1 and Ionic CLI >= 6.x
ionic capacitor sync <platform>ionic capacitor build <platform>For Capacitor version >= 3.1 use any of the following commands for instrumentation:
ionic capacitor run <platform>ionic capacitor sync <platform> && ionic capacitor open <platform>ionic capacitor build <platform>capacitor run <platform>capacitor sync <platform> && capacitor open <platform>Note:
Capacitor 3.1 added new hooks inside of their CLI. This allows us to instrument your application when using the ionic capacitor run or capacitor run commands.
Note: If both custom CLI arguments and custom configuration arguments are used, custom CLI arguments will be used.
module.exports = {
cordova : {
debug : false,
cspURL: "http://...",
cookieProxy: false,
jsAgentPath: "/path/to/jsAgent.txt",
gradlePath: "/path/to/build.gradle",
plistPath: "/path/to/Info.plist"
},
...
}
The default value is false. This property generates more log output and is sometimes necessary if you need to find the cause for a non-working plugin.
There is flag for updating the CSP (Content Security Policy). By default this value is set and the plugin will modify the CSP. The URL in the cspURL property will be placed into a CSP configuration (if available) in your index.html. This will allow/unblock connections to the Dynatrace server. If you don't want to use this feature, remove the cspURL property and the plugin will not modify the CSP configuration.
Issues with cookies are somehow a persistent companion in Ionic or Capacitor applications. Therefor we provide a proxy script which is wrapping the document.cookie API and are storing the cookies which are important for the Dynatrace Agents in the session storage. Of course this wrapping is still calling the original API.
Per default the cookie proxy is turned off. To make use of this proxy you need to enable it in the cordova configuration in the dynatrace.config.js:
module.exports = {
cordova : {
cookieProxy: true
},
...
}
The true in the cookieProxy property will insert a dt-cookie-proxy.js script in the index.html.
If you want to use a local script/text file that includes the JS Agent snippet downloaded from the WebUI or retrieved from using the Dynatrace API. Note that you can name the file whatever you want but the file and directory that you use needs to exist. The file downloaded from the WebUI will be named jsSnippet.txt by default.
If your top level build.gradle is not in the standard location or you want to use a specific build.gradle, you can specify a path to the desired file. We will assume that the other gradle file resides in /app/build.gradle. This will add all agent dependencies automatically for you and will update the configuration.
Tell the script where your Info.plist file is. The plist file is used for updating the configuration for the iOS agent.
module.exports = {
android : {
config : `...`
},
ios : {
config : `...`
}
}
The native configuration contain all the settings which are necessary for the Mobile Agent(s). You can find all the available properties in the documentation of the mobile agent.
The content of the ios.config property will be directly copied to the plist file. The content of the android.config property will be directly copied to the gradle file.
Here we list the configuration which is especially important if you are instrumenting a Hybrid application.
module.exports = {
android : {
config : `
dynatrace{
configurations {
defaultConfig{
autoStart{
...
}
hybridWebView{
enabled true
domains '.example.com', '.dynatrace.com'
}
}
}
}
`
},
ios : {
config : `
<key>DTXHybridApplication</key>
<true/>
<key>DTXSetCookiesForDomain</key>
<array>
<string>.example.com</string>
<string>.dynatrace.com</string>
</array>
`
}
}
DTXHybridApplication or hybridWebView.enabled : Set to true if you have a Hybrid application. The default value is false.
DTXSetCookiesForDomain or hybridWebView.domains : For hybrid applications using the JavaScript agent, cookies need to be set for each instrumented domain or server the application communicates with. You can specify domains, host or IP addresses. Domains or sub-domains must start with a dot. Separate the list elements with a comma.
Basically all needed properties for the JavaScript Agent are predefined by the downloadable dynatrace.config.js. There are three available properties:
url - Dynatrace API url to retrieve the JS agent script tag.mode - Values can be numbers 0-4 depending on what JavaScript Agent code snippet you want to use.allowanycert - Allows the plugin to ignore certificate issues when retrieving the JavaScript Agent.Using a specific mode can allow you during build to insert any of the available JavaScript agent code snippet options that are offered. Click here for more details on the options listed below:
jsInlineScriptjsTagCompletesyncCS (Default)asyncCSjsTagmodule.exports = {
js : {
url : `https://.../jsInlineScript/...`,
mode: 2
},
}
The above example will use mode option 2 and retrieve the synchronous code snippet of the JavaScript agent.
Note:
The default url value will be used if mode is not included in the dynatrace.config.js file or if mode exists and the value is not valid (i.e. not a number 0 through 4).
If you are having an issue retrieving the JavaScript Agent and see an error message relating to a certificate:
Could not retrieve the JavaScript Agent! - Could not retrieve agent optionsError: unable to verify the first certificate
You are able to bypass those errors at your OWN RISK by using allowanycert: true within the js property. This will ignore the fact that the SSL connection is not secure (e.g. because of invalid certificate) and will retrieve the JavaScript Agent snippet anyways. An example would look like this:
module.exports = {
js : {
url : `...`,
allowanycert: true
},
}
The JavaScript Agent interface will be provided by the JavaScript Agent, it can be used everywhere in your application by simply calling dtrum.
This gives you the possibility to instrument your code even further by adding manual instrumentation. If you like to know more about the manual instrumentation, have a look into the Dynatrace documentation.
To create a custom action that is called "MyButton tapped" you just need to use the following code below. The leaveAction will close the action again. It is possible to report errors for this action before closing, see next section Report Error.
"enterAction"(actionName: string, actionType: string, startTime?: number, sourceUrl?: string, sourceTitle?: string): number;
"leaveAction"(actionId: number, stopTime?: number, startTime?: number): void;
var action = dtrum.enterAction('simple action', 'click', null, 'http://whatever.com');
//do something here
dtrum.leaveAction(action);
For any open action you can report certain values. The following APIs are available on the Action:
reportError(error: Error | string, parentActionId?: number): void;
dtrum.reportError('Error: Hello World with AJAX!');
It is possible to identify a user and tag the current session with a name. You need to do the following call:
dtrum.identifyUser("User XY");
The above functionalities are only a small part of how you can use the API. If you want to know more you can visit the Settings in the WebUI.
If you are using typescript and want to use the manual instrumentation of the JSAgent you should install the typings. The typings are available under the package name @dynatrace/dtrum-api-types. External typings have the advantage that you can apply the version that is actually needed.
After installing the typings, they can be used by inserting the following line at the top of the file:
/// <reference types="@dynatrace/dtrum-api-types" />
or including the typings in your tsconfig.json file:
"files": [
"node_modules/@dynatrace/dtrum-api-types/dtrum.d.ts"
]
Be aware if you are using typescript you need to prefix your dtrum calls with window. as the typings file is made for all web applications.
This means a call to enterAction might look like this:
window.dtrum.enterAction(...)
or
let dtrum = window.dtrum;
dtrum.enterAction(...);
The Mobile Agent interface will be provided by the Mobile Agent, so it can be used everywhere in your application by simply calling dynatraceMobile.
In a hybrid scenario it is only possible for the mobile agent to end a session/visit. That's why we expose the endVisit function of the Mobile Agent.
The interface is available with the name dynatraceMobile (TypeScript definitions included). Calling dynatraceMobile.endVisit(successCallback, errorCallback) will end the session/visit. Example how this call looks like:
dynatraceMobile.endVisit(() => {
// Success
console.log("Visit was ended!");
}, () => {
// Error
console.log("Visit wasn't ended!");
});
You can also use the dtrum API directly to end the session:
dtrum.endSession(...);
Note: This API is not available in the typings file as we add it for convenience at runtime. If you are using typescript you might want to use dynatraceMobile.endSession(...).
Specifies if the user has to opt-in for being monitored. When enabled, you must specify the privacy setting. The following in configuration is needed to be able to use the API.
android: {
config: `
dynatrace {
configurations {
defaultConfig {
autoStart{
...
}
userOptIn true
}
}
}
`
}
ios: {
config: `
<key>DTXUserOptIn</key>
</true>
`
}
The privacy API methods allow you to dynamically change the data-collection level based on the individual preferences of your end users. Each end user can select from three data-privacy levels:
const enum DataCollectionLevel {
Off, Performance, UserBehavior
}
Here are some examples of how to use the API:
To check the current privacy options that are set:
dynatraceMobile.getUserPrivacyOptions((result) => {
// Successfully retrieve UserPrivacyOptions Object (i.e. result):
if(result) {
console.log(`Current Data Collection Level: ${result.dataCollectionLevel}`);
console.log(`Current Crash Reporting Value: ${result.crashReportingOptedIn}`);
}
}, (error) => {
// Error
console.log(`Error: ${error}`);
});
If you want to change the UserPrivacyOptions:
dynatraceMobile.applyUserPrivacyOptions(DataCollectionLevel.UserBehavior, true, () => {
console.log("Success");
}, () => {
console.log("Error");
});
To use the interface of the Mobile Agent directly you must specify the typing definition file in the tsconfig.json. Add the following block to the tsconfig.json:
"files": [
"node_modules/@dynatrace/cordova-plugin/typings/main.d.ts"
]
If "files" is already defined, just add the path to the already defined ones.
If you still don't get types in your .ts file, please insert the following snippet at the top of your .ts file:
/// <reference types="@dynatrace/cordova-plugin" />
Attention: This feature requires a instrumented webserver with version 1.211.x
Within Cordova it is also possible to use libraries that are executing native web requests. Some examples are: cordova-plugin-advanced-http, @ionic-native/http, cordova-plugin-okhttp or Mobile First by IBM.
Those libraries are not executing the request within the JS scope but will rather call a JavaScript<->Native interface and will trigger the request natively. This means that our JavaScript OneAgent will not be able to detect the request. Additionally a resulting user action can be dropped as well. To prevent this behaviour you need to manually apply headers to your request.
Before using this feature you need to turn of web request instrumentation on iOS. You can do this by adding the following configuration to the dynatrace.config.js:
ios: {
config: `
<key>DTXInstrumentWebRequestTiming</key>
<false/>
`
}
The following code will create a user action and a linked web request for you:
// Creation of your request. This might depend on the library you are using
MyTestRequest request = new MyTestRequest("MY_URL");
let interceptorUtils = dynatraceMobile.getNativeNetworkInterceptorUtils();
// Creation of a user action with automated dection of the user action - e.g. click on
let actionId = interceptorUtils.enterNativeRequestAction("MY_URL", "Name of HTTP Framework", "Optional Action Fallbackname");
// Adding headers to your request - Take a look into typings.d.ts for the format of DynatraceNativeRequestHeaders
request.addHeaders(interceptorUtils.getHeadersForNativeRequest(actionId));
// Sending the request and afterwards close the XhrAction
request.send();
interceptorUtils.leaveNativeRequestAction(actionId);
Attention: This feature requires a instrumented webserver with version 1.211.x running at your Mobile First server. Also, only version >= 8.x.x is supported.
If you are using IBM Mobile First you might take a look at our dynatraceMobile.getMobileFirstNetworkInterceptor(). This interceptor gives you the option to make the request going to Mobile First more visible. The Mobile First library is doing native mobile web requests, that's why those request can't be linked with user actions happening in the web part.
This interceptor is creating wrapping actions according to the user input that happened right before the request. By this modification you will be able to see the request in the user session linked with user actions.
Before using this feature you need to turn of web request instrumentation on iOS. You can do this by adding the following configuration to the dynatrace.config.js:
After Mobile First has been initialized make the following call:
dynatraceMobile.getMobileFirstNetworkInterceptor().enableInterceptor("https://serverurl.com");
if(dynatraceMobile.getMobileFirstNetworkInterceptor().isInterceptorEnabled()){
// Everything is setup correctly
}
It is also possible to deactivate the interceptor again. This will reset the modifications to the WLResourceRequest:
dynatraceMobile.getMobileFirstNetworkInterceptor().disableInterceptor();
if(!dynatraceMobile.getMobileFirstNetworkInterceptor().isInterceptorEnabled()){
// Interceptor was disabled
}
The Mobile agent usually sets cookies on the correct domain but the cordova-plugin-ionic-webview uses different domains. Those domains need to be added in the configuration. If you defined a custom hostname you need to take this into account as well for choosing the correct domain. A sample configuration which takes care about the default domain for cordova-plugin-ionic-webview looks like the following:
module.exports = {
android : {
config : `
dynatrace{
configurations {
defaultConfig{
autoStart{
...
}
hybridWebView{
enabled true
domains 'http://localhost', '.other-domain-you-want-to-specify.com'
}
}
}
}
`
},
ios : {
config : `
<DTXHybridApplication>true</DTXHybridApplication>
<DTXSetCookiesForDomain>ionic://localhost,.other-domain-you-want-to-specify.com</DTXSetCookiesForDomain>
`
}
}
The version can be used like in any other npm package. You just have to use the @ sign if you want to specify a certain version:
cordova plugin add @dynatrace/cordova-plugin@1.191.1 --save
This will download the version 1.191.1 of our plugin. In general we recommend you to always use the latest version.
Notes:
Our scripts assumes that the usual cordova project structure is given. The following arguments can be specified for our instrumentation script if the project structure is different.
--gradle=C:\MyCordovaAndroidProject\platforms\android\build.gradle - the location of the root build.gradle file. We will assume that the other gradle file resides in /app/build.gradle. This will add all agent dependencies automatically for you and will update the configuration.--plist=C:\MyCordovaIOSProject\platforms\ios\projectName\projectName-Info.plist - Tell the script where your Info.plist file is. The plist file is used for updating the configuration for the agent.--config=C:\SpecialFolderForDynatrace\dynatrace.config.js - Used for if you have not added your config file in the root folder of the Cordova project but somewhere else.--jsagent=C:\MyCordovaProject\scripts\jsSnippet.txt - If you want to use a local script/text file that includes the JS Agent snippet downloaded from the WebUI or retrieved from using the Dynatrace API. Note that you can name the file whatever you want but the file and directory that you use needs to exist. The file downloaded from the WebUI will be named jsSnippet.txt by default.Example:
cordova build android --config=C:\SpecialFolderForDynatrace\dynatrace.config.js
If you use Ionic make sure to use -- :
ionic cordova build android -- --config=C:\SpecialFolderForDynatrace\dynatrace.config.js
Additional Note:
For the local script file, we copy the specified file 1:1. It is best to download the JS agent snippet from the WebUI after selecting the desired JS agent format and clicking the Download link below the shown code snippet. To get to this page, go to your Web Application settings and then select Setup. The contents of the file will be copied 1:1 inside of the <head> tag of your Cordova applications platform based html files.
The differences to the old dynatrace-cordova-plugin are very small. There are three differences:
dynatrace.config to dynatrace.config.js. The new format can be downloaded via WebUI. If you had some custom settings in Android take a look in the detailed documentation. The format of iOS basically stayed the same.Updating Gradle only affects your Android build. To Update your project to Gradle 6 you have to modify one file in your Android folder.
ProjectFolder\android\cordova\lib\builders\ProjectBuilder.js Contains the following line:var distributionUrl = process.env['CORDOVA_ANDROID_GRADLE_DISTRIBUTION_URL'] || 'https\\://services.gradle.org/distributions/gradle-4.3.10-all.zip';
make sure you insert some other version like 6.1.1 here:
var distributionUrl = process.env['CORDOVA_ANDROID_GRADLE_DISTRIBUTION_URL'] || 'https\\://services.gradle.org/distributions/gradle-6.1.1-all.zip';
If you having trouble with this option you can always set the global process property CORDOVA_ANDROID_GRADLE_DISTRIBUTION_URL :
process.env['CORDOVA_ANDROID_GRADLE_DISTRIBUTION_URL'] = 'https\\://services.gradle.org/distributions/gradle-6.1.1-all.zip';
Because the Dynatrace Android agent now requires the MavenCentral repository, if either jcenter() or mavenCentral() is not added inside of ALL the repositories blocks via the top-level build.gradle, the build will fail.
Below is an example of what a basic top-level build.gradle file should look like after adding mavenCentral() to all repository blocks:
The location of the top-level build.gradle should be:
<rootOfProject>\platforms\android\build.gradle<rootOfProject>\android\build.gradleNote:
JCenter has noted its sunset on May 1st. Though, JCenter is still syncing with Maven Central so havingjcenter() in your build.gradle file without the use of mavenCentral() will retrieve the Dynatrace Android Gradle Plugin no problem.
If your application starts but you see no data (or the session is not merged), you probably need to dig deeper to find out why the OneAgents aren't sending any data. Opening up a support ticket is a great idea, but gathering logs first is even better. Please attach the mobile agent logs and the doctor dynatrace logs when opening a support ticket.
Add the following configuration snippet to your other configuration in dynatrace.config.js right under the autoStart block (the whole structure is visible, so you know where the config belongs):
android: {
config: `
dynatrace {
configurations {
defaultConfig {
autoStart{
...
}
debug.agentLogging true
}
}
}
`
}
Add the following configuration snippet to your other configuration in dynatrace.config.js (the whole structure is visible, so you know where the config belongs):
ios: {
config: `
<key>DTXLogLevel</key>
<string>ALL</string>
`
}
Doctor dynatrace is a tool that will analyze your project and determine if there is something wrong with your current configuration and will also provide information and requirements based on features that are available when using our plugin.
To run this tool, open terminal in the root directory of your project and run the following command: npm run doctorDynatrace
The results will be generated in the terminal window as well as stored in log folder located in the following directories:
C:\MyCordovaProject\plugins\dynatrace-cordova-plugin\logs\currentDoctorDynatrace.txtC:\MyCapacitorIOSProject\node_modules\@dynatrace\cordova-plugin\logs\currentDoctorDynatrace.txtThe log file will also include a list of your package.json dependencies and devDependencies which will be helpful when investigating possible issues that arise. Please add this doctor dynatrace log and mobile agent debug logs when opening a support ticket.
Note: The Dynatrace Android Gradle plugin is hosted on Maven Central. JCenter has noted it's sunset on May 1st so Maven Central is the primary source of the Dynatrace Android Gradle plugin.
Basically if you have problems with the plugin please have a look into the logs. They will tell you what went wrong. The logs can be found in the plugins folder of your Cordova project. There is a directory called "Logs".
--jsagent= custom parameter and download the full Javascript Agent and add the path to the downloaded JS Agent file to the custom parameter value - With this the plugin will not retrieve the JS agent and will use the one that is specified.iOS apps using ionic/capacitor have issues with passing cookies from the native layer to WkWebView layer. This issue is due to ionic and capacitor using their own default custom schemes (i.e. ionic:// or capacitor://) and the strict policy of Apple's WebKit not allowing cookies to be set in WkWebView with these custom schemes.
To resolve this issue and allow the Mobile and Web sessions to correlate properly, we inject the session information into local storage so that the JavaScript agent can use this info for the Web session it creates.
Requirements
When using Apple Pay with WKWebView, there are specific checks that WebKit (Apple) uses when loading the Apple Pay SDK. The Apple Pay SDK does not work when using script injection APIs which is what our iOS agent uses to properly communicate with the JS agent and to correlate the Mobile and Web User Sessions.
Workaround: You can set the DTXHybridApplication flag to false which will remove the injection of our script. This will also remove the correlation of the Mobile and Web Sessions.
1.255.1
1.253.5
1.247.0
dynatrace.config.js file1.233.1
1.229.0
1.227.6
1.225.1
1.225.0
1.219.1
1.217.1
1.213.0
1.205.0
1.201.0
1.192.0
1.191.2
FAQs
This plugin gives you the ability to use the Dynatrace instrumentation in your hybrid application (Cordova, Ionic, ..). It uses the Mobile Agent, the JavaScript Agent. The Mobile Agent will give you all device specific values containing lifecycle informat
The npm package @dynatrace/cordova-plugin receives a total of 243 weekly downloads. As such, @dynatrace/cordova-plugin popularity was classified as not popular.
We found that @dynatrace/cordova-plugin demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 0 open source maintainers collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
Research
A malicious npm package disguised as a WhatsApp client is exploiting authentication flows with a remote kill switch to exfiltrate data and destroy files.
Security News
PyPI now supports digital attestations, enhancing security and trust by allowing package maintainers to verify the authenticity of Python packages.
Security News
GitHub removed 27 malicious pull requests attempting to inject harmful code across multiple open source repositories, in another round of low-effort attacks.