Advanced configuration

Prevent duplicate application instances

When you launch an application multiple times, AppManager creates multiple application instances by default. You can use launch keys when starting applications to prevent multiple instances of the same application.

const launchOptions: AppLaunchOptions = {
    launchKey: 'Worklist-singleton'
};

// Returns existing instance if launch key matches a running application
const appHost = await appManager.launchApp("Worklist", launchOptions);

The launch key is a caller-generated value that prevents duplicate instances. When an application with a matching launch key is already running, AppManager returns the existing instance rather than creating a new instance and triggers onAppHostFocused event. This event could be leveraged to bring the application to focus when it's being launched again.

Dynamic application launch and management

AppManager provides onAppHostAdded and onAppHostRemoved notifications to indicate when a new application is launched or destroyed. These events could be leveraged to dynamically create and destroy iframes. For an example implementation, see Example implementation of dynamic application management with React section below.

Attach metadata to appHost

If you are managing iframes dynamically and want the storage of any application specific UI state or routing information in the AppHost, you can attach a custom JSON metadata at the time of application launch. This metadata is accessible in all event callbacks for application state management.

const launchOptions: AppLaunchOptions = {
    appManagerData: {
        // Custom JSON data
  }
};

const appHost = await appManager.launchApp("Worklist", launchOptions);

// Access the metadata
console.log("AppManagerData attached to AppHost", appHost.appManagerData);

Support Global Resiliency

Amazon Connect Global Resiliency features provide automatic failover between AWS Regions. When your Amazon Connect instance is configured for Global Resiliency, you should implement the following handlers to ensure agent workspace responsiveness to failover events:

// Handle pending failover events
connect.globalResiliency?.onFailoverPending(() => {
   // Destroy all applications before failover
    appManager.clearAll();
});

// Handle completed failover events
connect.globalResiliency?.onFailoverCompleted(() => {
    // Re-launch necessary applications following failover
    // Implementation depends on your state management approach
});

See Set up Amazon Connect Global Resiliency in the Amazon Connect Administrator Guide.