05 marzo 2019
Posted by Sumir Kataria, Software Engineering Lead & Jisha Abubaker, Product Manager
When we looked at the top problems faced by developers, we saw that doing background processing reliably and in a battery-friendly manner was a huge challenge. This meant that periodically fetching fresh content or uploading your logs was complex. Different versions of Android provided different tools for the job, each with their own API quirks. For example, listening for network or storage availability and automatically retrying your tasks involved a lot of work.
Our answer to these challenges was WorkManager. We introduced a preview of the Android Jetpack WorkManager library at Google I/O 2018 and have since iterated on it with additional features and bug fixes thanks to your valuable input.
The goal of WorkManager is to make background operations easy for you. WorkManager takes into account constraints like battery-optimization, storage, or network availability, and it only runs its tasks when the appropriate conditions are met. It also knows when to retry or reschedule your work--even if your device or app restarts.
We believe WorkManager is a friendly, approachable API that can take care of one of the most complex parts of Android for you so you can focus on the code that makes your app unique.
Here are some key features of WorkManager:
Watch and read below to learn when and how to use WorkManager to simplify managing background work in your apps:
WorkManager is best suited for tasks that can be deferred, but are still expected to run even if the application or device restarts (for example, syncing data periodically with a backend service and uploading logs or analytics data).
For tasks like sending an instant message that are required to run immediately or for tasks that are not required to run after the app exits, take a look at our background processing guide to learn which solution meets your needs.
To get started with the WorkManager API, add the WorkManager dependency available on Google's Maven repository in Java or Kotlin to your application's build.gradle
file:
dependencies { def work_version = 1.0.0 // Java implementation "android.arch.work:work-runtime:$work_version" // Kotlin KTX + coroutines implementation "android.arch.work:work-runtime-ktx:$work_version" }
Now, simply subclass a Worker and implement your background work with doWork() and enqueue it with WorkManager.
class MyWorker(ctx: Context, params: WorkerParameters) : Worker(ctx, params) { override fun doWork(): Result { //do the work you want done in the background here return Result.success() } } // optionally, add constraints like power, network availability val constraints: Constraints = Constraints.Builder() .setRequiresCharging(true) .setRequiredNetworkType(NetworkType.CONNECTED) .build() val myWork = OneTimeWorkRequestBuilder() .setConstraints(constraints).build()
WorkManager will now take care of running your task when it detects that your device is charging and the network is available.
WorkManager will leverage the right scheduling API under the hood: it uses JobScheduler API on Android 6.0+ (API 23+) and a combination of AlarmManager and BroadcastReceiver on previous versions.
It also seeks to ensure the best possible behavior so that it complies with system optimizations introduced in newer Android API versions to maximize battery and enforce good app behavior.
For example, WorkManager will schedule background work during the maintenance window for Android 6.0+ (API 23+) devices when the system is in Doze mode.
With WorkManager, you can easily add constraints like network availability or charging status. Your work will run when the constraints are met and automatically retried if they fail while running. For example, if your task requires network to be available, the task will be stopped when network is no longer available and retried later.
You can also monitor work status and retrieve work result using LiveData. This allows your UI to be notified when your task is completed.
In the event that your work fails, you can control how your work is retried by configuring how backoff is handled.
WorkManager is also able to reschedule your work, using a record of your work in its local database, if an application or device restart occurs.
We understand that each app has unique needs, and so do your tasks--even within the same app. WorkManager provides a simple yet highly flexible API surface to help configure your work and how it is run.
Take advantage of one-off scheduling with OneTimeWorkRequest or recurrent scheduling with PeriodicWorkRequest.
You can also chain your one time work requests to run in order or in parallel. If any work in the chain fails, WorkManager seeks to ensure that the remaining chain of work will not run. Read more about chaining work requests here.
If you require more flexibility over how WorkManager parallelizes and manages work, check out our advanced threading guide.
redBus, the largest online bus ticketing platform, shares their experience using WorkManager to simplify how they collect user feedback in their Android app:
"Feedback is critical to redBus as we expand into other countries. It often happens that a user gives critical feedback about a functionality within the redBus app but when the app tries to upload the feedback to backend servers, there might not be enough network coverage or battery.
WorkManager has simplified the way redBus app delivers information to it's backend servers. WorkManager library's capability to handle parameters like network connectivity, battery and use appropriate handlers like AlarmManager or JobScheduler has enabled us to concentrate on building business logics and offloading execution complexity to WorkManager."
- Dinesh Shanmugam
Android Lead, redBus.in
Check out our getting started guide and hands-on codelab to start using the WorkManager library for your background task needs.
We appreciate your feedback, including features you like and features you would like to see.
If you find a bug or issue, feel free to file an issue.