序
本文主要研究一下jvm的direct buffer统计
spring boot metrics
jvm.memory.used
代码语言：javascript复制{
"name": "jvm.memory.used",
"description": "The amount of used memory",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 137868592
}
],
"availableTags": [
{
"tag": "area",
"values": [
"heap",
"nonheap"
]
},
{
"tag": "id",
"values": [
"Compressed Class Space",
"PS Survivor Space",
"PS Old Gen",
"Metaspace",
"PS Eden Space",
"Code Cache"
]
}
]
}jvm.memory.used包括heap和nonheap两大类
heap
http://localhost:8080/actuator/metrics/jvm.memory.used?tag=area:heap
代码语言：javascript复制{
"name": "jvm.memory.used",
"description": "The amount of used memory",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 84724536
}
],
"availableTags": [
{
"tag": "id",
"values": [
"PS Eden Space",
"PS Survivor Space",
"PS Old Gen"
]
}
]
}heap的话根据具体的垃圾收集器类型有不同的区分
nonheap
http://localhost:8080/actuator/metrics/jvm.memory.used?tag=area:nonheap
代码语言：javascript复制{
"name": "jvm.memory.used",
"description": "The amount of used memory",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 54874872
}
],
"availableTags": [
{
"tag": "id",
"values": [
"Metaspace",
"Compressed Class Space",
"Code Cache"
]
}
]
}nonheap这里包括3个，分别是Metaspace、Compressed Class Space、Code Cache
jvm.buffer.memory.used
http://localhost:8080/actuator/metrics/jvm.buffer.memory.used?tag=id:direct
代码语言：javascript复制{
"name": "jvm.buffer.memory.used",
"description": "An estimate of the memory that the Java virtual machine is using for this buffer pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 81920
}
],
"availableTags": [
{
"tag": "id",
"values": [
"direct",
"mapped"
]
}
]
}jvm.buffer.memory.used包含了direct、mapped两大类
direct
http://localhost:8080/actuator/metrics/jvm.buffer.memory.used?tag=id:direct
代码语言：javascript复制{
"name": "jvm.buffer.memory.used",
"description": "An estimate of the memory that the Java virtual machine is using for this buffer pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 81920
}
],
"availableTags": []
}mapped
http://localhost:8080/actuator/metrics/jvm.buffer.memory.used?tag=id:mapped
代码语言：javascript复制{
"name": "jvm.buffer.memory.used",
"description": "An estimate of the memory that the Java virtual machine is using for this buffer pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 0
}
],
"availableTags": []
}jvm.buffer.count
http://localhost:8080/actuator/metrics/jvm.buffer.count
代码语言：javascript复制{
"name": "jvm.buffer.count",
"description": "An estimate of the number of buffers in the pool",
"baseUnit": "buffers",
"measurements": [
{
"statistic": "VALUE",
"value": 10
}
],
"availableTags": [
{
"tag": "id",
"values": [
"direct",
"mapped"
]
}
]
}jvm.buffer.count分direct和mapped两大类
direct
http://localhost:8080/actuator/metrics/jvm.buffer.count?tag=id:direct
代码语言：javascript复制{
"name": "jvm.buffer.count",
"description": "An estimate of the number of buffers in the pool",
"baseUnit": "buffers",
"measurements": [
{
"statistic": "VALUE",
"value": 10
}
],
"availableTags": []
}mapped
http://localhost:8080/actuator/metrics/jvm.buffer.count?tag=id:mapped
代码语言：javascript复制{
"name": "jvm.buffer.count",
"description": "An estimate of the number of buffers in the pool",
"baseUnit": "buffers",
"measurements": [
{
"statistic": "VALUE",
"value": 0
}
],
"availableTags": []
}jvm.buffer.total.capacity
http://localhost:8080/actuator/metrics/jvm.buffer.total.capacity
代码语言：javascript复制{
"name": "jvm.buffer.total.capacity",
"description": "An estimate of the total capacity of the buffers in this pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 81920
}
],
"availableTags": [
{
"tag": "id",
"values": [
"direct",
"mapped"
]
}
]
}jvm.buffer.total.capacity分direct和mapped两大类
direct
http://localhost:8080/actuator/metrics/jvm.buffer.total.capacity?tag=id:direct
代码语言：javascript复制{
"name": "jvm.buffer.total.capacity",
"description": "An estimate of the total capacity of the buffers in this pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 81920
}
],
"availableTags": []
}mapped
http://localhost:8080/actuator/metrics/jvm.buffer.total.capacity?tag=id:mapped
代码语言：javascript复制{
"name": "jvm.buffer.total.capacity",
"description": "An estimate of the total capacity of the buffers in this pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 0
}
],
"availableTags": []
}Native Memory Tracking
summary
代码语言：javascript复制jcmd 6878 VM.native_memory summary scale=MB
6878:
Native Memory Tracking:
Total: reserved=5625MB, committed=844MB
- Java Heap (reserved=4096MB, committed=577MB)
(mmap: reserved=4096MB, committed=577MB)
- Class (reserved=1066MB, committed=46MB)
(classes #7027)
(malloc=10MB #10535)
(mmap: reserved=1056MB, committed=37MB)
- Thread (reserved=36MB, committed=36MB)
(thread #37)
(stack: reserved=36MB, committed=36MB)
- Code (reserved=246MB, committed=15MB)
(malloc=2MB #3834)
(mmap: reserved=244MB, committed=12MB)
- GC (reserved=160MB, committed=148MB)
(malloc=10MB #220)
(mmap: reserved=150MB, committed=138MB)
- Internal (reserved=10MB, committed=10MB)
(malloc=10MB #10055)
- Symbol (reserved=9MB, committed=9MB)
(malloc=8MB #74319)
(arena=2MB #1)
- Native Memory Tracking (reserved=2MB, committed=2MB)
(tracking overhead=2MB)其中Internal部分包含了jvm中使用的directBuffer的大小
示例
代码语言：javascript复制 public void run(String... args) throws Exception {
// 分配一个256MB的直接缓冲区
ByteBuffer buffer = ByteBuffer.allocateDirect(256 * 1024 * 1024);
// 填充数据
Random random = new Random();
while (buffer.remaining() >= 4) {
buffer.putInt(random.nextInt());
}
System.out.println("Allocated direct buffer with capacity " buffer.capacity());
}VM.native_memory
代码语言：javascript复制jcmd 8077 VM.native_memory summary scale=MB
8077:
Native Memory Tracking:
Total: reserved=5881MB, committed=1099MB
- Java Heap (reserved=4096MB, committed=576MB)
(mmap: reserved=4096MB, committed=576MB)
- Class (reserved=1066MB, committed=46MB)
(classes #7028)
(malloc=10MB #10794)
(mmap: reserved=1056MB, committed=37MB)
- Thread (reserved=36MB, committed=36MB)
(thread #37)
(stack: reserved=36MB, committed=36MB)
- Code (reserved=246MB, committed=16MB)
(malloc=2MB #3889)
(mmap: reserved=244MB, committed=13MB)
- GC (reserved=160MB, committed=148MB)
(malloc=10MB #220)
(mmap: reserved=150MB, committed=138MB)
- Internal (reserved=266MB, committed=266MB)
(malloc=266MB #10055)
- Symbol (reserved=9MB, committed=9MB)
(malloc=8MB #74324)
(arena=2MB #1)
- Native Memory Tracking (reserved=2MB, committed=2MB)
(tracking overhead=2MB)可以看到Internal部分由之前的10MB增大到了266MB
jvm.buffer.memory.used
http://localhost:8080/actuator/metrics/jvm.buffer.memory.used?tag=id:direct
代码语言：javascript复制{
"name": "jvm.buffer.memory.used",
"description": "An estimate of the memory that the Java virtual machine is using for this buffer pool",
"baseUnit": "bytes",
"measurements": [
{
"statistic": "VALUE",
"value": 268476416
}
],
"availableTags": []
}可以看到jvm.buffer.memory.used的direct部分也变大了
小结
jvm的direct buffer可以通过springboot的jvm.buffer.memory.used?tag=id:direct来统计，也可以通过MNT的Internal部分反应出来。
doc
- 聊聊HotSpot VM的Native Memory Tracking
- 聊聊jvm的-XX:MaxDirectMemorySize
- 聊聊openjdk的BufferPoolMXBean
- 聊聊jvm的Code Cache
- 聊聊jvm的CompressedClassSpace
- 聊聊jvm的Stack Memory
- 聊聊jvm的StringTable及SymbolTable
