Groovy - JMX

JMX 是事實上的標準，用於監控任何與 Java 虛擬環境有關的應用程式。由於 Groovy 直接位於 Java 之上，因此 Groovy 可以利用大量為 Java 使用 JMX 所做的工作。

監視 JVM

可以使用 java.lang.management 中提供的標準類來執行 JVM 的監視。以下程式碼示例顯示瞭如何執行此操作。

import java.lang.management.*

def os = ManagementFactory.operatingSystemMXBean 
println """OPERATING SYSTEM: 
\tOS architecture = $os.arch 
\tOS name = $os.name 
\tOS version = $os.version 
\tOS processors = $os.availableProcessors 
""" 
 
def rt = ManagementFactory.runtimeMXBean 
println """RUNTIME: 
   \tRuntime name = $rt.name 
   \tRuntime spec name = $rt.specName 
   \tRuntime vendor = $rt.specVendor 
   \tRuntime spec version = $rt.specVersion 
   \tRuntime management spec version = $rt.managementSpecVersion 
   """ 

def mem = ManagementFactory.memoryMXBean 
def heapUsage = mem.heapMemoryUsage 
def nonHeapUsage = mem.nonHeapMemoryUsage 

println """MEMORY: 
   HEAP STORAGE: 
      \tMemory committed = $heapUsage.committed 
      \tMemory init = $heapUsage.init 
      \tMemory max = $heapUsage.max 
      \tMemory used = $heapUsage.used NON-HEAP STORAGE: 
      \tNon-heap memory committed = $nonHeapUsage.committed 
      \tNon-heap memory init = $nonHeapUsage.init 
      \tNon-heap memory max = $nonHeapUsage.max 
      \tNon-heap memory used = $nonHeapUsage.used 
   """
  
println "GARBAGE COLLECTION:" 
ManagementFactory.garbageCollectorMXBeans.each { gc ->
   println "\tname = $gc.name"
   println "\t\tcollection count = $gc.collectionCount"
   println "\t\tcollection time = $gc.collectionTime"
   String[] mpoolNames =   gc.memoryPoolNames
	
   mpoolNames.each { 
      mpoolName -> println "\t\tmpool name = $mpoolName"
   } 
}

當執行此程式碼時，輸出將根據執行程式碼的系統而有所不同。下面給出了輸出樣本。

OPERATING SYSTEM: 
   OS architecture = x86 
   OS name = Windows 7 
   OS version = 6.1 
   OS processors = 4
   
RUNTIME: 
   Runtime name = 5144@Babuli-PC 
   Runtime spec name = Java Virtual Machine Specification 
   Runtime vendor = Oracle Corporation 
   Runtime spec version = 1.7 
   Runtime management spec version = 1.2
   
MEMORY: 
   HEAP STORAGE: 
      Memory committed = 16252928 
      Memory init = 16777216 
      Memory max = 259522560 
      Memory used = 7355840
   
NON-HEAP STORAGE: 
   Non-heap memory committed = 37715968 
   Non-heap memory init = 35815424 
   Non-heap memory max = 123731968 
   Non-heap memory used = 18532232 
   
GARBAGE COLLECTION: 
   name = Copy 
   collection count = 15 
   collection time = 47 
   mpool name = Eden Space 
   mpool name = Survivor Space
		
   name = MarkSweepCompact 
      collection count = 0 
      collection time = 0 
		
      mpool name = Eden Space 
      mpool name = Survivor Space 
      mpool name = Tenured Gen 
      mpool name = Perm Gen 
      mpool name = Perm Gen [shared-ro] 
      mpool name = Perm Gen [shared-rw]

監視 Tomcat

為了監視 Tomcat，應在啟動 Tomcat 時設定以下引數 -

set JAVA_OPTS = -Dcom.sun.management.jmxremote 
Dcom.sun.management.jmxremote.port = 9004\
 
-Dcom.sun.management.jmxremote.authenticate=false 
Dcom.sun.management.jmxremote.ssl = false

以下程式碼使用 JMX 在執行的 Tomcat 中發現可用的 MBean，確定哪些是 Web 模組並提取每個 Web 模組的處理時間。

import groovy.swing.SwingBuilder
  
import javax.management.ObjectName 
import javax.management.remote.JMXConnectorFactory as JmxFactory 
import javax.management.remote.JMXServiceURL as JmxUrl 
import javax.swing.WindowConstants as WC 
 
import org.jfree.chart.ChartFactory 
import org.jfree.data.category.DefaultCategoryDataset as Dataset 
import org.jfree.chart.plot.PlotOrientation as Orientation 
 
def serverUrl = 'service:jmx:rmi:///jndi/rmi://:9004/jmxrmi' 
def server = JmxFactory.connect(new JmxUrl(serverUrl)).MBeanServerConnection 
def serverInfo = new GroovyMBean(server, 'Catalina:type = Server').serverInfo 
println "Connected to: $serverInfo" 
 
def query = new ObjectName('Catalina:*') 
String[] allNames = server.queryNames(query, null) 

def modules = allNames.findAll { name -> 
   name.contains('j2eeType=WebModule') 
}.collect{ new GroovyMBean(server, it) }
  
println "Found ${modules.size()} web modules. Processing ..." 
def dataset = new Dataset() 
 
modules.each { m ->
   println m.name()
   dataset.addValue m.processingTime, 0, m.path 
}