-XX:OnError="pmap %p" 
 --> JVM에서 Fatal Error가 발생하면 Process ID에 대해 pmap 명령을 수행한다.

-XX:OnOutOfMemoryError="pmap %p" 
 --> JVM에서 Fatal Error가 발생하면 Process ID에 대해 pmap 명령을 수행한다.

[GC [DefNew: 64575K->959K(64576K), 0.0457646 secs] 196016K->133633K(261184K), 0.0459067 secs]]

• GC 전의 Young Generation Usage = 64M, Young Generation Size = 64M
• GC 전의 Total Heap Usage = 196M, Total Heap Size = 260M
• GC 후의 Young Generation Usage = 9.5M
• GC 후의 Total Heap Usage = 133M
• Minor GC 소요 시간 = 0.0457646 초

111.042: [GC 111.042: [DefNew: 8128K->8128K(8128K), 0.0000505 secs]
111.042: [Tenured: 18154K->2311K(24576K), 0.1290354 secs] 26282K->2311K(32704K), 0.1293306 secs]

• GC 전의 Tenured Generation Usage = 18M, Tenured Generation Size = 24M
• GC 후의 Tenured Generation Usage = 2.3M
• Major GC 소요시간 = 0.12초

0.548403: [GC {Heap before GC invocations=1:
Heap
par new generation total 18432K, used 12826K [0xf2800000, 0xf4000000, 0xf4000000]
eden space 12288K, 99% used<1> [0xf2800000, 0xf33ff840, 0xf3400000]
from space 6144K, 8% used<2> [0xf3a00000, 0xf3a87360, 0xf4000000]
to space 6144K, 0% used<3> [0xf3400000, 0xf3400000, 0xf3a00000]
concurrent mark-sweep generation total 40960K, used 195K<4>[0xf4000000, 0xf6800000, 0xf6800000]
CompactibleFreeListSpace space 40960K, 0% used [0xf4000000, 0xf6800000]
concurrent-mark-sweep perm gen total 4096K, used 1158K<5> [0xf6800000, 0xf6c00000, 0xfa800000]
CompactibleFreeListSpace space 4096K, 28% used [0xf6800000, 0xf6c00000]
0.549364: [ParNew: 12826K<6>->1086K<7>(18432K<8>), 0.02798039 secs] 13022K->1282K(59392K)
Heap after GC invocations=2:
Heap
par new generation total 18432K, used 1086K [0xf2800000, 0xf4000000, 0xf4000000]
eden space 12288K, 0% used<10> [0xf2800000, 0xf2800000, 0xf3400000]
from space 6144K, 17% used<11> [0xf3400000, 0xf350fbc0, 0xf3a00000]
to space 6144K, 0% used<12> [0xf3a00000, 0xf3a00000, 0xf4000000]
concurrent mark-sweep generation total 40960K, used 195K<13> [0xf4000000, 0xf6800000, 0xf6800000]
CompactibleFreeListSpace space 40960K, 0% used [0xf4000000, 0xf6800000]
concurrent-mark-sweep perm gen total 4096K, used 1158K<14> [0xf6800000, 0xf6c00000, 0xfa800000]
CompactibleFreeListSpace space 4096K, 28% used [0xf6800000, 0xf6c00000]
} , 0.0297669 secs]

• all - The default if just -memorycheck is used. Enables checking of all allocated and freed blocks on every free and allocate call. This check of the heap is the most thorough and should cause the JVM to exit on nearly all memory-related problems very soon after they are caused. This option has the greatest impact on performance.
• quick - Enables block padding only. Used to detect basic heap corruption. Pads every allocated block with sentinel bytes, which are verified on every allocate and free. Block padding is faster than the default of checking every block, but is not as effective.
• nofree - Keeps a list of already used blocks instead of freeing memory. This list is checked, along with currently allocated blocks, for memory corruption on every allocation and deallocation. Use this option to detect a dangling pointer (a pointer that is "dereferenced" after its target memory is freed). This option cannot be reliably used with long-running applications (such as WAS), because "freed" memory is never reused or released by the JVM.
• failat=<number of allocations> - Causes memory allocation to fail (return NULL) after <number of allocations>. Setting <number of allocations> to 13 will cause the 14th allocation to return NULL. Deallocations are not counted. Use this option to ensure that JVM code reliably handles allocation failures. This option is useful for checking allocation site behavior rather than setting a specific allocation limit.
• skipto=<number of allocations> - Causes the program to check only on allocations that occur after <number of allocations>. Deallocations are not counted. Used to speed up JVM startup when early allocations are not causing the memory problem. As a rough estimate, the JVM performs 250+ allocations during startup.
• callsite=<number of allocations> - Prints callsite information every <number of allocations>. Deallocations are not counted. Callsite information is presented in a table with separate information for each callsite. Statistics include the number and size of allocation and free requests since the last report, and the number of the allocation request responsible for the largest allocation from each site. Callsites are presented as sourcefile:linenumber for C code and assembly function name for assembler code. Callsites that do not provide callsite information are accumulated into an "unknown" entry.
• zero - Newly allocated blocks are set 0 instead of being filled with the 0xE7E7xxxxxxxxE7E7 pattern. Setting to 0 helps you to determine whether a callsite is expecting zeroed memory (in which case the allocation request should be followed by memset(pointer, 0, size)).

• class - Class Loading 정보를 Standard Err에 출력한다. 출력 예는 아래와 같다.

class load: java/io/FilePermission
class load: java/io/FilePermissionCollection
class load: java/security/AllPermission
...
class load: test
class load: test from: file:/C:/Documents/Java_Test/GC%20dump/

• dynload - Class Loading에 대한 매우 상세한 정보를 제공한다. 클래스명, 클래스 크기, 로딩 시간등의 정보를 포함한다. 출력 예는 아래와 같다.

<  Class size 6594; ROM size 7056; debug size 0> 
<  Read time 128 usec; Load time 126 usec; Translate time 222 usec>
<Loaded java/security/BasicPermissionCollection from c:\IBM\WebSphere\AppServer\java\jre\lib\core.jar>
<  Class size 4143; ROM size 3264; debug size 0>
<  Read time 103 usec; Load time 81 usec; Translate time 117 usec>
<Loaded java/security/Principal from c:\IBM\WebSphere\AppServer\java\jre\lib\core.jar>
<  Class size 239; ROM size 248; debug size 0>
<  Read time 44 usec; Load time 23 usec; Translate time 20 usec>
<Loaded test>
<  Class size 370; ROM size 448; debug size 0>
<  Read time 0 usec; Load time 28 usec; Translate time 39 usec>

• gc - GC 작업에 대한 정보를 제공한다. 자세한 내용은 GC Dump를 참조한다.
• jni - JNI 호출에 대한 정보를 제공한다. 출력 예는 아래와 같다.

<JNI ReleaseStringChars: buffer=41EC45B8>
<JNI GetStaticMethodID: gc_dump.main ([Ljava/lang/String;)V>
<JNI GetMethodID: java/lang/reflect/Method.getModifiers ()I>
<JNI GetMethodID: java/lang/String.<init> ([B)V>
<JNI FindClass: java/lang/Object>
<JNI GetMethodID: java/lang/Object.finalize ()V>
<JNI FindClass: java/lang/ref/Reference>
<JNI GetMethodID: java/lang/ref/Reference.enqueueImpl ()Z>

• sizes - Memory 사용과 관련된 설정값을 출력한다. 출력 예는 아래와 같다.

 -Xmca32K              RAM class segment increment
 -Xmco128K            ROM class segment increment
 -Xmns0K                initial new space size
 -Xmnx0K                maximum new space size
 -Xms4M                 initial memory size
 -Xmos4M               initial old space size
 -Xmox1047608K     maximum old space size
 -Xmx1047608K       memory maximum
 -Xmr16K               remembered set size
 -Xmso32K             OS thread stack size
 -Xiss2K                java thread stack initial size
 -Xssi16K              java thread stack increment
 -Xss256K             java thread stack maximum size
 -Xscmx16M          shared class cache size

• stacks - Thread 별로 Java/C Stack의 사용 크기를 출력한다. 출력 예는 아래와 같다.

JVMVERB000I Verbose stack: "Thread-1" used 188/3756 bytes on Java/C stacks
JVMVERB000I Verbose stack: "Thread-2" used 516/3756 bytes on Java/C stacks
JVMVERB000I Verbose stack: "main" used 1368/0 bytes on Java/C stacks
JVMVERB000I Verbose stack: "Finalizer thread" used 456/2308 bytes on Java/C stacks
JVMVERB000I Verbose stack: "Gc Slave Thread" used 232/3060 bytes on Java/C stacks

• optthruput: Throughput Collector를 사용한다. 처리량(Throughput)을 최적화할 목적으로 사용되며, Default Collector이다.
• optavgpause: Response Time Collector를 사용한다. 응답시간(Response Time)을 최적화할 목적으로 사용된다. GC에 의한 Pause Time을 최소하기 위해 Concurrent Mark/Sweep 작업을 수행한다. Throughput Collector에 비해 처리량으로 다소(5~10%) 떨어진다.
• gencon: Concurrent Generational Collector를 사용한다. IBM JDK 1.5에서 추가되었다. Sun Hotspot JVM의 CMS Collector와 거의 동일한 방식으로 동작하다.
• subpool: Subpool Collector를 사용한다.

...
TR_GlobalRegisterAllocator::perform()
TR_OptimizerImpl::performOptimization()
TR_OptimizerImpl::performOptimization()
TR_OptimizerImpl::optimize()
...

# Inlining을 비활성화한다.
-Xjit:disableInlining
-Xjit:{java/lang/Math.max(II)I}(disableInlining)

# 특정 메소드를 Optimization에서 제외한다.
-Xjit:exclude={java/lang/Math.max(II)I} ...

count=<n>
   <n>번 째 수행에서 Method를 JIT 컴파일한다. JIT 문제를 추적할 때는 "0"의 값을 사용함으로써 보다 빨리 컴파일이 이루어지도록 한다.    
optlevel=[noOpt | cold | warm | hot | veryHot | scorching]
   [비최적화 ~ 최고 최적화]까지 JIT 컴파일에 의한 최적화 레벨을 지적한다.
verbose
   JIT 컴파일 과정에서 사용된 정보와 컴파일 과정을 출력한다.

JIT type: Testarossa (Full)
JIT options specified:
    verbose
options in effect:
    bcount=250
    catchSamplingSizeThreshold=1100
    classLoadPhaseInterval=500
    classLoadPhaseThreshold=155
    code=512 (KB)
    codepad=0 (KB)
    codetotal=0 (KB)
    count=1000
    ...
    stack=256
    target=ia32-win32
    verbose=1
 
+ (warm) java/lang/Double.doubleToRawLongBits(D)J @ 0x41630014-0x41630030
+ (warm) java/lang/System.getEncoding(I)Ljava/lang/String; @ 0x41630054-0x41630145
+ (warm) java/lang/String.hashCode()I @ 0x4163017C-0x4163024A
+ (warm) java/util/HashMap.put(Ljava/lang/Object;Ljava/lang/Object;)Ljava/lang/Object; @ 0x4163027C-0x416304AF
+ (warm) java/util/Locale.toLowerCase(Ljava/lang/String;)Ljava/lang/String; @ 0x416304DC-0x416307FF
...
+ (warm) java/io/FileOutputStream.writeBytes([BIILjava/io/FileDescriptor;)V @ 0x41636C34-0x41636D45|-