以hbase为例：在hbase的配置文件路径下，设置了GC log输出路径/app/hbase-config/hbase-env.sh
export HBASE_OPTS="-Xmx16384m -Xms16384m -Xmn8192m -XX:PermSize=160M -XX:MaxPermSize=160M -XX:+UseConcMarkSweepGC -XX:+UseParNewGC -XX:CMSInitiatingOccupancyFraction=70 -XX:+UseCMSCompactAtFullCollection -XX:CMSFullGCsBeforeCompaction=3 -XX:+CMSParallelRemarkEnabled -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -XX:+PrintGCApplicationConcurrentTime -XX:+PrintGCApplicationStoppedTime -XX:PrintFLSStatistics=1 -Xloggc:/app/hbase/logs/hbase_gc.log $HBASE_OPTS"可以查询到log的相关信息
#jstat -gccause 79751 1000 1000 //获取GC相关指标，表示每1000毫秒查询一次79751垃圾收集状况。
Usage: jstat -help|-options
jstat -<option> [-t] [-h<lines>] <vmid> [<interval> [<count>]]
参数解释：
Options — 选项，我们一般使用 -gcutil 查看gc情况
vmid — VM的进程号，即当前运行的java进程号
interval– 间隔时间，单位为秒或者毫秒
count — 打印次数，如果缺省则打印无数次
S0 S1 E O P YGC YGCT FGC FGCT GCT LGCC GCC
0.00 8.98 11.16 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 12.79 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 14.94 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 16.04 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 17.98 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 20.34 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 21.83 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
0.00 8.98 22.64 69.83 12.33 156 50.788 16 280.103 330.891 unknown GCCause No GC
E列代表
O E–>O 新生代转移到老代，Ｏ积累到80%就会触发full gc
具体意思如下：
E（表示，Eden），代表这台服务器的新生代Eden区使用了11.16的空间，
S0 S1 两个Survivor区(S0,S1表示Survivor0、Survivor1),Survivor0里面是空的，Survivor1占了8.98%
老年代（O，表示old) 和永久带（P，表示Permanent）分别使用了69.83% 和12.33%的空间。
程序运行以来共发生Minor GC(YGC ,表示young gc)156次，总耗时50.788秒；
发生Fulle GC （FGC，表示full GC ）16次，full GC总耗时（FGCT，表示full gc time）为330.891秒，总的GC总耗时（GCT，表示GC Time）为330.891秒
LGCC Cause of last Garbage Collection.
GCC Cause of current Garbage Collection.
Java 6 JDK输出如下：
[hadoop@0321 logs]$ jmap -heap 29877
Attaching to process ID 29877, please wait...
Debugger attached successfully.
Server compiler detected.
JVM version is 20.45-b01
using parallel threads in the new generation.
using thread-local object allocation.
Concurrent Mark-Sweep GC
Heap Configuration:
MinHeapFreeRatio = 40
MaxHeapFreeRatio = 70
MaxHeapSize = 21474836480 (20480.0MB)
NewSize = 4294967296 (4096.0MB)
MaxNewSize = 4294967296 (4096.0MB)
OldSize = 5439488 (5.1875MB)
NewRatio = 2
SurvivorRatio = 8
PermSize = 67108864 (64.0MB)
MaxPermSize = 67108864 (64.0MB)
Heap Usage:
New Generation (Eden + 1 Survivor Space):
capacity = 3865509888 (3686.4375MB)
used = 481048576 (458.763671875MB)
free = 3384461312 (3227.673828125MB)
12.444634470949257% used
Eden Space:
capacity = 3436052480 (3276.875MB)
used = 481048576 (458.763671875MB)
free = 2955003904 (2818.111328125MB)
14.00003576196834% used
From Space:
capacity = 429457408 (409.5625MB)
used = 0 (0.0MB)
free = 429457408 (409.5625MB)
0.0% used
To Space:
capacity = 429457408 (409.5625MB)
used = 0 (0.0MB)
free = 429457408 (409.5625MB)
0.0% used
concurrent mark-sweep generation:
capacity = 12884901888 (12288.0MB)
used = 0 (0.0MB)
free = 12884901888 (12288.0MB)
0.0% used
Perm Generation:
capacity = 67108864 (64.0MB)
used = 21153816 (20.173851013183594MB)
free = 45955048 (43.826148986816406MB)
31.521642208099365% used- Eden from to 3个加起来为Young 区
[hbase@0321 ~]$ jps
10424 Jps
8229 HRegionServer
[hbase@0321 ~]$ jmap -heap 8229
Attaching to process ID 8229, please wait…
Debugger attached successfully.
Server compiler detected.
JVM version is 24.71-b01
using thread-local object allocation.
Garbage-First (G1) GC with 6 thread(s)
Heap Configuration:
MinHeapFreeRatio = 40
MaxHeapFreeRatio = 70
MaxHeapSize = 25769803776 (24576.0MB)
NewSize = 1363144 (1.2999954223632812MB)
MaxNewSize = 17592186044415 MB
OldSize = 5452592 (5.1999969482421875MB)
NewRatio = 2
SurvivorRatio = 8
PermSize = 16777216 (16.0MB)
MaxPermSize = 83886080 (80.0MB)
G1HeapRegionSize = 8388608 (8.0MB)
Heap Usage:
G1 Heap:
regions = 3072
capacity = 25769803776 (24576.0MB)
used = 20409483264 (19464.0MB)
free = 5360320512 (5112.0MB)
79.19921875% used
G1 Young Generation:
Eden Space:
regions = 89
capacity = 1182793728 (1128.0MB)
used = 746586112 (712.0MB)
free = 436207616 (416.0MB)
63.12056737588652% used
Survivor Space:
regions = 20
capacity = 167772160 (160.0MB)
used = 167772160 (160.0MB)
free = 0 (0.0MB)
100.0% used
G1 Old Generation:
regions = 2324
capacity = 24419237888 (23288.0MB)
used = 19495124992 (18592.0MB)
free = 4924112896 (4696.0MB)
79.83510821023704% used
Perm Generation:
capacity = 50331648 (48.0MB)
used = 42303832 (40.344078063964844MB)
free = 8027816 (7.655921936035156MB)
84.05016263326009% used
G1 GC笔记：
关于最大gc停顿时间
在做minor gc和mixed gc时，收集器会在内存中维系一个remembered set，这个set包含了heap中所有对象的引用，用以确定哪些可以被回收。
每次gc时，g1会通过一个预测模型来计算每个region进行回收的时间，从而从中选出停顿时间在MaxGCPauseMillis之内的region进行垃圾回收
关于mixed gc
mixed gc是一种既回收young区，也回收old区的垃圾回收方式，它触发的条件是-XX:InitiatingHeapOccupancyPercent，mixed gc的目的就是为了延迟full gc的产生
同理YGC指的是回收young的垃圾回收方式
Full GC 指的是回收Old去的垃圾回收方式，因为old区比较大，导致GC时间比较长，并且在GC期间java进程停止对外相应，并且自身也不对外响应，导致regionserver无法向zookeeper注册心跳信息，超过zookeeper （ <name>zookeeper.session.timeout</name>） session时间，就会被zookeeper置为dead。
Hbase G1参数调整：
待修改的参数，修改GC的参数需要重启集群，安排在下次重启。
-XX:ConcGCThreads=12并发标记的执行线程数 =================== 测试数据ConcGCThreads 不能大于-XX:ParallelGCThreads，否则会报错
-XX:InitiatingHeapOccupancyPercent=60 堆占用了多少的时候就触发GC，默认为45
-XX:ParallelGCThreads=12 30
[hbase@0321 ~]$ jstat -gccapacity 61540
NGCMN NGCMX NGC S0C S1C EC OGCMN OGCMX OGC OC PGCMN PGCMX PGC PC YGC FGC
0.0 52428800.0 2490368.0 0.0 294912.0 2195456.0 0.0 52428800.0 44695552.0 44695552.0 16384.0 81920.0 49152.0 49152.0 2 0
GC 常见命令：
[hbase@ 0321 ~]$ jmap -heap 61540
Attaching to process ID 61540, please wait...
Debugger attached successfully.
Server compiler detected.
JVM version is 24.71-b01
using thread-local object allocation.
Garbage-First (G1) GC with 30 thread(s)
Heap Configuration:
MinHeapFreeRatio = 40
MaxHeapFreeRatio = 70
MaxHeapSize = 53687091200 (51200.0MB)
NewSize = 1363144 (1.2999954223632812MB)
MaxNewSize = 17592186044415 MB
OldSize = 5452592 (5.1999969482421875MB)
NewRatio = 2
SurvivorRatio = 8
PermSize = 16777216 (16.0MB)
MaxPermSize = 83886080 (80.0MB)
G1HeapRegionSize = 16777216 (16.0MB)
Heap Usage:
G1 Heap:
regions = 2880
capacity = 48318382080 (46080.0MB)
used = 2519304864 (2402.5963439941406MB)
free = 45799077216 (43677.40365600586MB)
5.213967760403951% used
G1 Young Generation:
Eden Space:
regions = 7
capacity = 2248146944 (2144.0MB)
used = 117440512 (112.0MB)
free = 2130706432 (2032.0MB)
5.223880597014926% used
Survivor Space:
regions = 18
capacity = 301989888 (288.0MB)
used = 301989888 (288.0MB)
free = 0 (0.0MB)
100.0% used
G1 Old Generation:
regions = 126
capacity = 45768245248 (43648.0MB)
used = 2099874464 (2002.5963439941406MB)
free = 43668370784 (41645.40365600586MB)
4.588059805705051% used
Perm Generation:
capacity = 50331648 (48.0MB)
used = 42610464 (40.636505126953125MB)
free = 7721184 (7.363494873046875MB)
84.65938568115234% used
13018 interned Strings occupying 1388240 bytes.[hbase@0321 ~]$ jstat -gccause -h 10 61540 1000 1000
S0 S1 E O P YGC YGCT FGC FGCT GCT LGCC GCC
0.00 100.00 92.54 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 92.54 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 92.54 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 93.28 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 93.28 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 93.28 20.26 85.06 16 1.794 0 0.000 1.794 G1 Evacuation Pause No GC
0.00 100.00 1.42 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 1.42 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
S0 S1 E O P YGC YGCT FGC FGCT GCT LGCC GCC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 3.55 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
S0 S1 E O P YGC YGCT FGC FGCT GCT LGCC GCC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
0.00 100.00 4.26 21.24 85.06 17 1.877 0 0.000 1.877 G1 Evacuation Pause No GC
#jstat -gcutil 27912 1s 50
原创文章，作者：奋斗，如若转载，请注明出处：https://blog.ytso.com/194413.html