fragmentPagerAdapter中item为几个不同的fragment(不是同一个数据源，不单是数据序号问题，三个独立的fragment，布局也不同)，有内存泄漏问题，我是通过保存了几个fragment

求大佬支援

关于android防止内存的dump，被折磨惨了/(ㄒoㄒ)/~~

好不容易找来找去，找到个解决方法，但是用不来。。。

内存溢出了 应该是这两个for循环出了问题 请问怎么编写更好一点

ArrayList<Entry>

麻烦来位同学帮我看看，Leaks工具不太会使，教教我，排查了两天了🤪🤪

这是 log

LeakCanary: ​
    ====================================
    HEAP ANALYSIS RESULT
    ====================================
    0 APPLICATION LEAKS
    
    References underlined with "~~~" are likely causes.
    Learn more at https://squ.re/leaks.
    ====================================
    1 LIBRARY LEAKS
    
    A Library Leak is a leak caused by a known bug in 3rd party code that you do not have control over.
    See https://square.github.io/leakcanary/fundamentals-how-leakcanary-works/#4-categorizing-leaks
    
    Leak pattern: instance field android.app.ActivityThread$ActivityClientRecord#nextIdle
    Description: Android AOSP sometimes keeps a reference to a destroyed activity as a nextIdle client record in the
    android.app.ActivityThread.mActivities map. Not sure what's going on there, input welcome.
    113742 bytes retained by leaking objects
    Displaying only 1 leak trace out of 2 with the same signature
    Signature: b29426c08253f2b2cd5cea6cc80e6fb5b5dc7b3
    ┬───
    │ GC Root: System class
    │
    ├─ android.app.ActivityThread class
    │    Leaking: NO (a class is never leaking)
    │    ↓ static ActivityThread.sCurrentActivityThread
    │                            ~~~~~~~~~~~~~~~~~~~~~~
    ├─ android.app.ActivityThread instance
    │    Leaking: UNKNOWN
    │    Retaining 517853 bytes in 3389 objects
    │    mInitialApplication instance of cn.cqautotest.sunnybeach.app.AppApplication
    │    mSystemContext instance of android.app.ContextImpl, not wrapping known Android context
    │    ↓ ActivityThread.mActivities
    │                     ~~~~~~~~~~~
    ├─ android.util.ArrayMap instance
    │    Leaking: UNKNOWN
    │    Retaining 73 bytes in 3 objects
    │    ↓ ArrayMap.mArray
    │               ~~~~~~
    ├─ java.lang.Object[] array
    │    Leaking: UNKNOWN
    │    Retaining 32 bytes in 1 objects
    │    ↓ Object[].[1]
    │               ~~~
    ├─ android.app.ActivityThread$ActivityClientRecord instance
    │    Leaking: UNKNOWN
    │    Retaining 502263 bytes in 3100 objects
    │    Library leak match: instance field android.app.ActivityThread$ActivityClientRecord#nextIdle
    │    activity instance of cn.cqautotest.sunnybeach.ui.activity.HomeActivity with mDestroyed = false
    │    ↓ ActivityThread$ActivityClientRecord.nextIdle
    │                                          ~~~~~~~~
    ├─ android.app.ActivityThread$ActivityClientRecord instance
    │    Leaking: UNKNOWN
    │    Retaining 499675 bytes in 3057 objects
    │    activity instance of cn.cqautotest.sunnybeach.ui.activity.SettingActivity with mDestroyed = true
    │    ↓ ActivityThread$ActivityClientRecord.activity
    │                                          ~~~~~~~~
    ╰→ cn.cqautotest.sunnybeach.ui.activity.SettingActivity instance
    ​     Leaking: YES (ObjectWatcher was watching this because cn.cqautotest.sunnybeach.ui.activity.SettingActivity
    ​     received Activity#onDestroy() callback and Activity#mDestroyed is true)
    ​     Retaining 104677 bytes in 933 objects
    ​     key = b72b98e5-e83c-4fe2-97b4-345cddb1e261
    ​     watchDurationMillis = 5935
    ​     retainedDurationMillis = 926
    ​     mApplication instance of cn.cqautotest.sunnybeach.app.AppApplication
    ​     mBase instance of androidx.appcompat.view.ContextThemeWrapper, not wrapping known Android context
    
    ====================================

泄漏日志截图

先贴个图

场景：部署博客门户网站后，只要有 发送请求，哪怕只有一个，服务器CPU占用100% 。如上图所示，node 和 docker-proxy两个占用了全部cpu。

服务器配置：1核4G 云硬盘100G。

请问这是什么情况=.=! 拉大锯的部署教程里使用的服务器好像是2G的，都没出现过这样的情况。