性能调优基础篇之内存测试常识

博客主页：http://blog.csdn.net/minna_d

性能调优时结果通常需要保证结果的一致性才有意义。

为了保证结果的一致性，必须了解到已下命令， 屏蔽机器、缓存等一系列外围因素

1. df -h 查看磁盘挂载点及内存使用情况

[root@jjy ~]# df -h
文件系统        容量  已用  可用 已用% 挂载点
/dev/sda3        97G   63G   30G   69% /
tmpfs           3.9G     0  3.9G    0% /dev/shm
/dev/sdb        459G   36G  400G    9% /notSolidHd

2. 测试磁盘读写速度
[root@jjy ~]# hdparm -t /dev/sda3
/dev/sda3:
 Timing buffered disk reads: 2
448 MB in  3.03 seconds = 147.85 MB/sec

3.查看物理内存使用情况

[root@jjy ~]# free -m
             total       used       free     shared    buffers     cached
Mem:          7842       1008       6833          0          2        426
-/+ buffers/cache:        579       7263
Swap:         3906          0       3906

 4. 清空系统缓存

xxxxxxxxxxxxxx: free -m
             total       used       free     shared    buffers     cached
Mem:          8192       8162         29          0        139       2519
-/+ buffers/cache:       5503       2688
Swap:         4094       2642       1451

echo 1  > /proc/sys/vm/drop_caches

xxxxxxxxxxxxxx: free -m

             total       used       free     shared    buffers     cached
Mem:          8192       6105       2086          0          1        592
-/+ buffers/cache:       5511       2680
Swap:         4094       2642       1451

0 – 不释放

1 – 释放页缓存，清空最近访问过的文件页面缓存

2 – 释放dentries和inodes，清空文件节点缓存和目录项缓存

3 – 释放所有缓存，清空1和2所有内容的缓存

5. top 命令， 查看内存， cpu使用情况， 不过对于java程序， 注意res 与 virt  项， 代表的是最大内存容量和常驻内存量，

并不是GC之后最低内存量，内存优化之后通常看的不是top ,而是fullgc 之后内存量或者用下一个jmap命令。

top - 11:06:44 up 964 days, 20:12,  1 user,  load average: 0.06, 0.04, 0.00
Tasks: 143 total,   1 running, 142 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.2%us,  0.0%sy,  0.0%ni, 99.5%id,  0.0%wa,  0.0%hi,  0.0%si,  0.3%st
Mem:   8388608k total,  6252204k used,  2136404k free,     2688k buffers
Swap:  4192924k total,  2706288k used,  1486636k free,   618396k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                                                                                          
21874 tomcat    18   0 8832m 5.0g  10m S  0.0 62.6  40:10.09 java                                                                                                                             
25593 postgres  16   0  124m  26m  25m S  0.0  0.3   2080:58 postgres                                                                                                                         
 4477 postgres  15   0  126m  21m  20m S  0.0  0.3   0:00.79 postgres  

6. jmap -heap pid 查看jvm内存结构分布

[root@jjy ~/q/data/qtypeahead]# jmap -heap 1742
Attaching to process ID 1742, please wait...
Debugger attached successfully.
Server compiler detected.
JVM version is 24.45-b08

using thread-local object allocation.
Parallel GC with 4 thread(s)

Heap Configuration:
   MinHeapFreeRatio = 40
   MaxHeapFreeRatio = 70
   MaxHeapSize      = 7864320000 (7500.0MB)
   NewSize          = 1310720 (1.25MB)
   MaxNewSize       = 17592186044415 MB
   OldSize          = 5439488 (5.1875MB)
   NewRatio         = 2
   SurvivorRatio    = 8
   PermSize         = 21757952 (20.75MB)
   MaxPermSize      = 85983232 (82.0MB)
   G1HeapRegionSize = 0 (0.0MB)

Heap Usage:
PS Young Generation
Eden Space:
   capacity = 874512384 (834.0MB)
   used     = 874512384 (834.0MB)
   free     = 0 (0.0MB)
   100.0% used
From Space:
   capacity = 873463808 (833.0MB)
   used     = 594060480 (566.5402221679688MB)
   free     = 279403328 (266.45977783203125MB)
   68.01203147274535% used
To Space:
   capacity = 873463808 (833.0MB)
   used     = 289505280 (276.09375MB)
   free     = 583958528 (556.90625MB)
   33.14450780312125% used
PS Old Generation
   capacity = 5242880000 (5000.0MB)
   used     = 4480736448 (4273.163269042969MB)
   free     = 762143552 (726.8367309570312MB)
   85.46326538085937% used
PS Perm Generation
   capacity = 22544384 (21.5MB)
   used     = 22447888 (21.407974243164062MB)
   free     = 96496 (0.0920257568359375MB)
   99.57197322401889% used

7.动态展示内存变化情况jconsole pid& ,我不喜欢截图，请自己尝试

8. jmap -histo:live pid 查看程序稳定之后内存

[root@jjy ~/q/data/qtypeahead]# jmap -histo:live 1742 > histo_live
[root@jjy ~/q/data/qtypeahead]# head histo_live  && tail histo_live

 num     #instances         #bytes  class name
----------------------------------------------
   1:      31395319     1336843920  [C
   2:      24818078      992723120  java.util.TreeMap$Entry
   3:      31394031      753456744  java.lang.String
   4:      24816294      595591056  com.qunar.hotel.qtypeahead.datastructure.QElement
   5:       7193675      345296400  java.util.HashMap
   6:       7151274      343261152  com.qunar.hotel.qtypeahead.datastructure.QElementMin
   7:       2902593      233245400  [Ljava.util.HashMap$Entry;
2427:             1             16  com.sun.jmx.remote.internal.ArrayNotificationBuffer$BroadcasterQuery
2428:             1             16  [Ljava.lang.management.MonitorInfo;
2429:             1             16  [Lorg.codehaus.jackson.map.deser.ValueInstantiators;
2430:             1             16  com.sun.org.apache.xerces.internal.dom.CharacterDataImpl$1
2431:             1             16  org.apache.http.impl.DefaultHttpResponseFactory
2432:             1             16  sun.reflect.GeneratedSerializationConstructorAccessor17
2433:             1             16  sun.nio.cs.CharsetMapping$4
2434:             1             16  com.qunar.base.meerkat.http.SSLVerifier
2435:             1             16  com.google.gson.internal.bind.TypeAdapters$12
Total     147825569     5032606128

9. 数据库也有自己的缓存， 同一条sql执行时间第一次比二次慢

hs=# \timing
Timing is on.
hs=# select count(*) from xx;
  count  
---------
 7403942
(1 row)

Time: 6542.208 ms
hs=# select count(*) from xx;
  count  
---------
 7403942
(1 row)

Time: 2022.684 ms

所以测试sql时最好使用explain analyze减少差异

hs=# explain analyze select count(*) from xx;
                                                                   QUERY PLAN                                                                   
------------------------------------------------------------------------------------------------------------------------------------------------
 Aggregate  (cost=231919.62..231919.64 rows=1 width=0) (actual time=21282.771..21282.773 rows=1 loops=1)
   ->  Seq Scan on qt_hotel_log_analyse_result  (cost=0.00..213329.30 rows=7436130 width=0) (actual time=1.081..11185.959 rows=7403942 loops=1)
 Total runtime: 21282.839 ms
(3 rows)

Time: 21302.382 ms
hs=# explain analyze select count(*) from xx;
                                                                   QUERY PLAN                                                                   
------------------------------------------------------------------------------------------------------------------------------------------------
 Aggregate  (cost=231919.62..231919.64 rows=1 width=0) (actual time=20388.296..20388.297 rows=1 loops=1)
   ->  Seq Scan on qt_hotel_log_analyse_result  (cost=0.00..213329.30 rows=7436130 width=0) (actual time=0.056..10702.320 rows=7403942 loops=1)
 Total runtime: 20388.353 ms
(3 rows)

Time: 20389.865 ms