Android 性能优化（23）＊性能工具之「Heap Viewer, Memory Monitor, Allocation Tracker」Memory Profilers

Memory Profilers

　　In this document

　　You should also read

　　Heap Viewer, Memory Monitor, and Allocation Tracker are complementary tools that visualize the memory your app uses.

Use the Memory Monitor Tool to find out whether undesirable garbage collection (GC) event patterns might be causing your performance problems.
```
Memory Monitor Tool 可分析出哪里有内存问题。
```
Run Heap Viewer to identify object types that get or stay allocated unexpectedely or unecessarily.
```
Heap Viewer 标识出哪个对象是不必要的分配。
```
Use Allocation Tracker to identify where in your code the problem might be.
```
Allocation Tracker 定位问题代码
```

Shows available and used memory as they both change over time for your app in a graph, with dips indicating Garbage Collection (GC) events.

Provides a quick way to test whether app slowness might be related to excessive GC events.
```
可测试是否由于过多的GC导致应用变慢。
```
Provides a quick way to see whether app crashes might be related to running out of memory.
```
检查是否由于内存不足导致应用崩溃。
```
Runs live on your application and updates about once a second.
Helps quickly identify the potential of memory leaks.
```
标识潜在的内存泄漏。
```
Helps identify GC patterns for your application and determine whether they are healthy and what you expect.
```
标识GC运行是否健康。
```
Is easy to use, and it is easy to interpret what you see.
However, Memory Monitor does not tell you which objects are the cause of your problems, or where in your code it might be.
```
它的缺点是并不能指出是哪个对象有问题，哪句代码有问题。
```

　　　　　　Figure 1. Memory Monitor.

Shows snapshots of a number of objects allocated by type.
Samples data every time a garbage collection event occurs naturally or triggered by you.
```
采集每次自动或手动GC的对象。
```
Helps identify which object types might be involved in memory leaks.
```
标识出哪些对象可能与内存泄漏有关。
```
However, you have to look for changes over time yourself by tracking what's happening in the graph.

　　　　　　　　　　Figure 2. Heap Viewer.

Shows when and where your code allocates object types, their size, allocating thread, and stack traces.
```
记录内存分配信息。
```
Helps recognize memory churn through recurring allocation/deallocation patterns.
```
识别重复无效的内存分配。
```
Can be used in combination with Heap Viewer to track down memory leaks. For example, if you see a bitmap object resident on the heap, you can find its allocation location with Allocation Tracker.
```
帮助Heap Viewer, Memory Monitor 分析内存泄漏。
```
However, it takes time and experience to learn to interpret the output from this tool.