xref: /frameworks/base/docs/html/topic/performance/memory.jd

page.title=Manage Your App's Memory
page.tags=ram,low memory,OutOfMemoryError,onTrimMemory

@jd:body

<div id="qv-wrapper">
<div id="qv">

<h2>In this document</h2>
<ol>
  <li><a href="#monitor">Monitor Available Memory and Memory Usage</a>
    <ul>
      <li><a href="#AnalyzeRam">Tools for analyzing RAM usage</a></li>
      <li><a href="#release">Release memory in response to events</a></li>
      <li><a href="#CheckHowMuchMemory">Check how much memory you should use</a></li>
    </ul>
  </li>
  <li><a href="#code">Use More Efficient Code Constructs</a>
    <ul>
      <li><a href="#Services">Use services sparingly</a></li>
      <li><a href="#DataContainers">Use optimized data containers</a></li>
      <li><a href="#Abstractions">Be careful with code abstractions</a></li>
      <li><a href="#NanoProto">Use nano protobufs for serialized data</a></li>
      <li><a href="#churn">Avoid memory churn</a></li>
    </ul>
  </li>
  <li><a href="#remove">Remove Memory-Intensive Resources and Libraries</a>
    <ul>
      <li><a href="#reduce">Reduce overall APK size</a></li>
      <li><a href="#DependencyInjection">Avoid dependency injection frameworks</a></li>
      <li><a href="#ExternalLibs">Be careful about using external libraries</a></li>
    </ul>
  </li>
</ol>
<h2>See Also</h2>
<ul>
  <li><a href="{@docRoot}training/articles/memory-overview.html">Overview of Android Memory Management</a>
  </li>
  <li><a href="{@docRoot}studio/profile/investigate-ram.html">Investigating Your RAM Usage</a>
  </li>
  <li><a href="{@docRoot}topic/performance/reduce-apk-size.html">Reduce APK Size</a></li>
</ul>

</div>
</div>

<!-- INTRO #################################################### -->

<p>
  Random-access memory (RAM) is a valuable
  resource in any software development environment, but
  it's even more valuable on a mobile operating system
  where physical memory is often constrained.
  Although both the Android Runtime (ART) and Dalvik virtual machine perform
  routine garbage collection, this does not mean you can ignore
  when and where your app allocates and releases memory.
  You still need to avoid
  introducing memory leaks, usually caused by holding onto
  object references in static member variables, and
  release any {@link java.lang.ref.Reference} objects at the appropriate
  time as defined by
  lifecycle callbacks.
</p>

<p>
  This page explains how you can
  proactively reduce memory usage within your app.
  For more information about general
  practices to clean up your resources when programming in Java,
  refer to other books or online
  documentation about managing resource references.
  If youre looking for information about how to
  analyze memory in a running app, read
  <a href="#AnalyzeRam">Tools for analyzing RAM usage</a>.
  For more detailed information about how the Android Runtime and Dalvik
  virtual machine manage memory, see the
  <a href="{@docRoot}training/articles/memory-overview.html">Overview of Android Memory Management</a>.
</p>

<!-- Section 1 #################################################### -->

<h2 id="monitor">Monitor Available Memory and Memory Usage</h2>

<p>
  The Android framework, Android Studio, and Android SDK
  can help you analyze and adjust your app's memory usage.
  The Android framework
  exposes several APIs that allow your app to reduce its memory usage
  dynamically during runtime. Android Studio and the Android SDK
  contain several tools  that allow you to investigate how your
  app uses memory.
</p>

<!-- Section 1.1 #################################################### -->

<h3 id="AnalyzeRam">Tools for analyzing RAM usage</h3>

<p>
  Before you can fix the memory usage problems in your app, you first need
  to find them. Android Studio and the Android SDK include several tools
  for analyzing memory usage in your app:
</p>

<ol>
  <li>
    The Device Monitor has a Dalvik Debug Monitor Server (DDMS) tool that allows
    you to inspect memory allocation within your app process.
    You can use this information to understand how your
    app uses memory overall. For example, you can force a garbage collection
    event and then view the types of objects that remain in memory. You can
    use this information to identify operations or actions within your app
    that allocate or leave excessive amounts of objects in memory.

    <p>For more information about how to use the DDMS tool, see
      <a href="/studio/profile/ddms.html">Using DDMS</a>.
    </p>
  </li>

  <li>
    The Memory Monitor in Android Studio shows you how your app allocates
    memory over the course of a single session.
    The tool shows a graph of available
    and allocated Java memory over time, including garbage collection events.
    You can also initiate garbage collection events and take a snapshot of
    the Java heap while your app runs. The output from the Memory Monitor tool
    can help you identify points when your app experiences excessive garbage
    collection events, leading to app slowness.
    <p>
      For more information about how to use Memory Monitor tool, see
      <a href="{@docRoot}tools/debugging/debugging-memory.html#ViewHeap">Viewing Heap Updates</a>.
    </p>
  </li>

  <li>
    Garbage collection events also show up in the Traceview viewer. Traceview
    allows you to view trace log files as both a timeline and as a profile
    of what happened within a method. You can use this tool to determine
    what code was executing when a garbage collection event occurred.
    <p>
      For more information about how to use the Traceview viewer, see
      <a href="https://developer.android.com/studio/profile/traceview.html">Profiling with Traceview and dmtracedump</a>.
    </p>
  </li>

  <li>
    The Allocation Tracker tool in Android Studio gives you a detailed look
    at how your app allocates memory.
    The Allocation Tracker records an app's memory allocations and lists
    all allocated objects within the profiling snapshot. You can use this
    tool to track down parts of your code that allocate too many objects.

    <p>
      For more information about how to use the Allocation Tracker tool, see
      <a href="{docRoot}studio/profile/allocation-tracker-walkthru.html">Allocation Tracker Walkthrough</a>.
    </p>
  </li>

</ol>

<!-- Section 1.2 #################################################### -->

<h3 id="release">Release memory in response to events</h3>

<p>
  An Android device can run with varying amounts of free memory
  depending on the physical amount of RAM on the device and how the user
  operates it. The system broadcasts signals to indicate when it is under
  memory pressure, and apps should listen for these signals and adjust
  their memory usage as appropriate.
</p>

</p>
  You can use the {@link android.content.ComponentCallbacks2} API
  to listen for these signals and then adjust your memory
  usage in response to app lifecycle
  or device events. The
  {@link android.content.ComponentCallbacks2#onTrimMemory onTrimMemory()}
  method allows your app to listen for memory related events when the app runs
  in the foreground (is visible) and when it runs in the background.
</p>

<p>
  To listen for these events, implement the {@link
  android.content.ComponentCallbacks2#onTrimMemory onTrimMemory()}
  callback in your {@link android.app.Activity}
  classes, as shown in the following code snippet.
</p>

<pre class="prettyprint">
import android.content.ComponentCallbacks2;
// Other import statements ...

public class MainActivity extends AppCompatActivity
    implements ComponentCallbacks2 {

    // Other activity code ...

    /**
     * Release memory when the UI becomes hidden or when system resources become low.
     * @param level the memory-related event that was raised.
     */
    public void onTrimMemory(int level) {

        // Determine which lifecycle or system event was raised.
        switch (level) {

            case ComponentCallbacks2.TRIM_MEMORY_UI_HIDDEN:

                /*
                   Release any UI objects that currently hold memory.

                   The user interface has moved to the background.
                */

                break;

            case ComponentCallbacks2.TRIM_MEMORY_RUNNING_MODERATE:
            case ComponentCallbacks2.TRIM_MEMORY_RUNNING_LOW:
            case ComponentCallbacks2.TRIM_MEMORY_RUNNING_CRITICAL:

                /*
                   Release any memory that your app doesn't need to run.

                   The device is running low on memory while the app is running.
                   The event raised indicates the severity of the memory-related event.
                   If the event is TRIM_MEMORY_RUNNING_CRITICAL, then the system will
                   begin killing background processes.
                */

                break;

            case ComponentCallbacks2.TRIM_MEMORY_BACKGROUND:
            case ComponentCallbacks2.TRIM_MEMORY_MODERATE:
            case ComponentCallbacks2.TRIM_MEMORY_COMPLETE:

                /*
                   Release as much memory as the process can.

                   The app is on the LRU list and the system is running low on memory.
                   The event raised indicates where the app sits within the LRU list.
                   If the event is TRIM_MEMORY_COMPLETE, the process will be one of
                   the first to be terminated.
                */

                break;

            default:
                /*
                  Release any non-critical data structures.

                  The app received an unrecognized memory level value
                  from the system. Treat this as a generic low-memory message.
                */
                break;
        }
    }
}
</pre>

<p>
  The
  {@link android.content.ComponentCallbacks2#onTrimMemory onTrimMemory()}
  callback was added in Android 4.0 (API level 14). For earlier versions,
  you can use the
  {@link android.content.ComponentCallbacks#onLowMemory()}
  callback as a fallback for older versions, which is roughly equivalent to the
  {@link android.content.ComponentCallbacks2#TRIM_MEMORY_COMPLETE} event.
</p>

<!-- Section 1.3 #################################################### -->

<h3 id="CheckHowMuchMemory">Check how much memory you should use</h3>

<p>
  To allow multiple running processes, Android sets a hard limit
  on the heap size alloted for each app. The exact heap size limit varies
  between devices based on how much RAM the device
  has available overall. If your app has reached the heap capacity and
  tries to allocate more
  memory, the system throws an {@link java.lang.OutOfMemoryError}.
</p>

<p>
  To avoid running out of memory, you can to query the system to determine
  how much heap space you have available on the current device.
  You can query the system for this figure by calling
  {@link android.app.ActivityManager#getMemoryInfo(android.app.ActivityManager.MemoryInfo) getMemoryInfo()}.
  This returns an
  {@link android.app.ActivityManager.MemoryInfo } object that provides
  information about the device's
  current memory status, including available memory, total memory, and
  the memory threshold&mdash;the memory level below which the system begins
  to kill processes. The
  {@link android.app.ActivityManager.MemoryInfo } class also exposes a simple
  boolean field,
  {@link android.app.ActivityManager.MemoryInfo#lowMemory }
  that tells you whether the device is running low on memory.
</p>

<p>
  The following code snippet shows an example of how you can use the
  {@link android.app.ActivityManager#getMemoryInfo(android.app.ActivityManager.MemoryInfo) getMemoryInfo()}.
  method in your application.
</p>

<pre class="prettyprint">
public void doSomethingMemoryIntensive() {

    // Before doing something that requires a lot of memory,
    // check to see whether the device is in a low memory state.
    ActivityManager.MemoryInfo memoryInfo = getAvailableMemory();

    if (!memoryInfo.lowMemory) {
        // Do memory intensive work ...
    }
}

// Get a MemoryInfo object for the device's current memory status.
private ActivityManager.MemoryInfo getAvailableMemory() {
    ActivityManager activityManager = (ActivityManager) this.getSystemService(ACTIVITY_SERVICE);
    ActivityManager.MemoryInfo memoryInfo = new ActivityManager.MemoryInfo();
    activityManager.getMemoryInfo(memoryInfo);
    return memoryInfo;
}
</pre>

<!-- Section 2 #################################################### -->

<h2 id="code">Use More Memory-Efficient Code Constructs</h2>

<p>
  Some Android features, Java classes, and code constructs tend to
  use more memory than others. You can minimize how
  much memory your app uses by choosing more efficient alternatives in
  your code.
</p>

<!-- Section 2.1 #################################################### -->

<h3 id="Services">Use services sparingly</h3>

<p>
  Leaving a service running when its not needed is
  <strong>one of the worst memory-management
  mistakes</strong> an Android app can make. If your app needs a
  <a href="{@docRoot}guide/components/services.html">service</a>
  to perform work in the background, do not keep it running unless
  it needs to run a job. Remember to stop your service when it has completed
  its task. Otherwise, you can inadvertently cause a memory leak.
</p>

<p>
  When you start a service, the system prefers to always keep the process
  for that service running. This behavior
  makes services processes very expensive
  because the RAM used by a service remains unavailable to other processes.
  This reduces the number of cached processes that the system can keep in
  the LRU cache, making app switching less efficient. It can even lead to
  thrashing in the system when memory is tight and the system cant
  maintain enough processes to host all the services currently running.
</p>

<p>
  You should generally avoid use of persistent services because of
  the on-going demands they place on available memory. Instead, we
  recommend that you use an alternative implementation
  such as {@link android.app.job.JobScheduler}. For more information about
  how to use {@link android.app.job.JobScheduler} to schedule background
  processes, see
  <a href="/topic/performance/background-optimization.html">Background Optimizations</a>.
<p>
  If you must use a service, the
  best way to limit the lifespan of your service is to use an {@link
  android.app.IntentService}, which finishes
  itself as soon as it's done handling the intent that started it.
  For more information, read
  <a href="{@docRoot}training/run-background-service/index.html">Running in a Background Service</a>.
</p>

<!-- Section 2.2 #################################################### -->

<h3 id="DataContainers">Use optimized data containers</h3>

<p>
  Some of the classes provided by the programming language are not optimized for
  use on mobile devices. For example, the generic
  {@link java.util.HashMap} implementation can be quite memory
  inefficient because it needs a separate entry object for every mapping.
</p>

<p>
  The Android framework includes several optimized data containers, including
  {@link android.util.SparseArray}, {@link android.util.SparseBooleanArray},
  and {@link android.support.v4.util.LongSparseArray}.
  For example, the {@link android.util.SparseArray} classes are more
  efficient because they avoid the system's need to
  <acronym title="Automatic conversion from primitive types to object classes (such as int to Integer)">autobox</acronym>
  the key and sometimes value (which creates yet another object or
  two per entry).
</p>

<p>
  If necessary, you can always switch to raw arrays for a really lean data
  structure.
</p>

<!-- Section 2.3 #################################################### -->

<h3 id="Abstractions">Be careful with code abstractions</h3>

<p>
  Developers often use abstractions simply as a good programming practice,
  because abstractions can improve code flexibility and maintenance.
  However, abstractions come at a significant cost:
  generally they require a fair amount more code that
  needs to be executed, requiring more time and
  more RAM for that code to be mapped into memory.
  So if your abstractions aren't supplying a
  significant benefit, you should avoid them.
</p>

<p>
  For example, enums often require more than twice as much memory as static
  constants. You should strictly avoid using enums on Android.
</p>

<!-- Section 2.4 #################################################### -->

<h3 id="NanoProto">Use nano protobufs for serialized data</h3>

<p>
  <a href="https://developers.google.com/protocol-buffers/docs/overview">Protocol buffers</a>
  are a language-neutral, platform-neutral, extensible mechanism
  designed by Google for serializing structured data&mdash;similar to XML, but
  smaller, faster, and simpler. If you decide to use
  protobufs for your data, you should always use nano protobufs in your
  client-side code. Regular protobufs generate extremely verbose code, which
  can cause many kinds of problems in your app such as
  increased RAM use, significant APK size increase, and slower execution.
</p>

<p>
  For more information, see the "Nano version" section in the
  <a href="https://android.googlesource.com/platform/external/protobuf/+/master/java/README.txt"
class="external-link">protobuf readme</a>.
</p>

<!-- Section 2.5 #################################################### -->

<h3 id="churn">Avoid memory churn</h3>

<p>
  As mentioned previously, garbage collections events don't normally affect
  your app's performance. However, many garbage collection events that occur
  over a short period of time can quickly eat up your frame time. The more time
  that the system spends on garbage collection, the less time it has to do
  other stuff like rendering or streaming audio.
</p>

<p>
  Often, <em>memory churn</em> can cause a large number of
  garbage collection events to occur. In practice, memory churn describes the
  number of allocated temporary objects that occur in a given amount of time.
</p>

<p>
  For example, you might allocate multiple temporary objects within a
  <code>for</code> loop. Or you might create new
  {@link android.graphics.Paint} or {@link android.graphics.Bitmap}
  objects inside the
  {@link android.view.View#onDraw(android.graphics.Canvas) onDraw()}
  function of a view.
  In both cases, the app creates a lot of objects quickly at high volume.
  These can quickly consume all the available memory in the young generation,
  forcing a garbage collection event to occur.
</p>

<p>
  Of course, you need to find the places in your code where
  the memory churn is high before you can fix them. Use the tools discussed in
  <a href="#AnalyzeRam">Analyze your RAM usage</a>
</p>

<p>
  Once you identify the problem areas in your code, try to reduce the number of
  allocations within performance critical areas. Consider moving things out of
  inner loops or perhaps moving them into a
  <a href="https://en.wikipedia.org/wiki/Factory_method_pattern" class="external-link">Factory</a>
  based allocation structure.
</p>

<!-- Section 3 #################################################### -->

<h2 id="remove">Remove Memory-Intensive Resources and Libraries</h2>

<p>
  Some resources and libraries within your code can gobble up memory without
  you knowing it. Overall size of your APK, including third-party libraries
  or embedded resources, can affect how much memory your app consumes. You can
  improve your app's memory consumption by removing any redundant, unnecessary,
  or bloated components, resources, or libraries from your code.
</p>

<!-- Section 3.1 #################################################### -->

<h3 id="reduce">Reduce overall APK size</h3>

<p>
  You can significantly reduce your app's memory usage by reducing the overall
  size of your app. Bitmap size, resources, animation frames, and third-party
  libraries can all contribute to the size of your APK.
  Android Studio and the Android SDK provide multiple tools
  to help you reduce the size of your resources and external dependencies.
</p>

<p>
  For more information about how to reduce your overall APK size, see
  <a href="{@docRoot}topic/performance/reduce-apk-size.html">Reduce APK Size</a>.
</p>

<!-- Section 3.2 #################################################### -->

<h3 id="DependencyInjection">Use caution with dependency injection frameworks</h3>

<p>
  Dependency injection framework such as
  <a href="https://code.google.com/p/google-guice/" class="external-link">Guice</a>
  or
  <a href="https://github.com/roboguice/roboguice" class="external-link">RoboGuice</a>
  can simplify the code you write and provide an adaptive environment
  that's useful for testing and other configuration changes. However, dependency
  frameworks aren't always optimized for mobile devices.
</p>

<p>
  For example, these frameworks tend to initialize processes by
  scanning your code for annotations. This which can require significant
  amounts of your code to be mapped into RAM unnecessarily. The system
  allocates these mapped pages into clean memory so Android can drop them; yet
  that can't happen until the pages have remained in memory for a long period
  of time.
 </p>

<p>
  If you need to use a dependency injection framework in your app, consider
  using
  <a class="external-link" href="http://google.github.io/dagger/">Dagger</a>
  instead. For example, Dagger does not use reflection to scan your app's code.
  Dagger's strict implementation means that it can be used in Android apps
  without needlessly increasing memory usage.
</p>

<!-- Section 3.3 #################################################### -->

<h3 id="ExternalLibs">Be careful about using external libraries</h3>

<p>
  External library code is often not written for mobile environments and
  can be inefficient when used
  for work on a mobile client. When you decide to use an
  external library, you may need to optimize that library for mobile devices.
  Plan for that work up-front and analyze the library in terms of code size and
  RAM footprint before deciding to use it at all.
</p>

<p>
  Even some mobile-optimized libraries can cause problems due to differing
  implementations. For example, one library may use nano protobufs
  while another uses micro protobufs, resulting in two different protobuf
  implementations in your app. This can happen with different
  implementations of logging, analytics, image loading frameworks,
  caching, and many other things you don't expect.
</p>

<p>
  Although <a href="{@docRoot}tools/help/proguard.html">ProGuard</a> can
  help to remove APIs and resources with the right flags, it can't remove a
  library's large internal dependencies. The features that you want in these
  libraries may require lower-level dependencies. This becomes especially
  problematic when you use an {@link android.app.Activity } subclass from a
  library (which will tend to have wide swaths of dependencies),
  when libraries use reflection (which is common and means you need to spend a
  lot of time manually tweaking ProGuard to get it to work), and so on.
</p>

<p>
  Also avoid using a shared library for just one or two features out of dozens.
  You don't want to pull in a large amount of code and overhead that
  you don't even use. When you consider whether to use a library, look for
  an implementation that strongly matches what you need. Otherwise, you might
  decide to create your own implementation.
</p>