最近一直没有更新这部分的内容,会利用五一时间完成vcenter这一个系列。
这里先给大家一本关于vijava开发的书,比较实用。
地址:http://pan.baidu.com/s/1gfkl9mj。密码:t1y3
有网友评论要数据存储读写速度及延迟之后时间的实现。今天就先介绍关于java实现实时监控vcenter状态的内容,包括CPU、内存、网络、存储等。
在看这篇前如果还是新手,建议先看下我的第一篇 http://www.cnblogs.com/xiaodige/p/6721517.html(vijava基本连接和数据中心信息获取)
先说一下我之前是怎么发现,CPU、内存、网络、存储等可以监控的内容。我用vsphere client连接vcenter查看客户端所能监控的属性.
建议大家在实现功能前,先看下官方文档关于性能监控的知识。贴上连接vcenter6.0的官方文档连接。http://pubs.vmware.com/vsphere-60/index.jsp
比如要监控虚拟机的性能信息:首先得保证虚拟机开机,选择“性能选项”,点击“图标选项”,这样就能看到它所支持的性能监控信息。下面贴图:
下图左边就是该对象具体能监控的信息,右边就是具体性能,使用率啊等等。
做性能监控这块,如果不知道具体能监控对象的哪些属性就打开客户端看看。下面就贴上具体的java实现代码。
package com.iking.vmware.performance; import java.util.ArrayList; import java.util.Calendar; import java.util.Date; import java.util.HashMap; import java.util.List; import java.util.Map; import com.iking.vmware.bean.PerformanceManage; import com.iking.vmware.bean.VsphereConst; import com.iking.vmware.connection.ConnectedVimServiceBase; import com.iking.vmware.vim25.PerfCounterInfo; import com.iking.vmware.vim25.PerfEntityMetric; import com.iking.vmware.vim25.PerfEntityMetricBase; import com.iking.vmware.vim25.PerfMetricId; import com.iking.vmware.vim25.PerfMetricIntSeries; import com.iking.vmware.vim25.PerfMetricSeries; import com.iking.vmware.vim25.PerfQuerySpec; import com.iking.vmware.vim25.PerfSampleInfo; import com.iking.vmware.vim25.mo.Folder; import com.iking.vmware.vim25.mo.HostSystem; import com.iking.vmware.vim25.mo.InventoryNavigator; import com.iking.vmware.vim25.mo.ManagedEntity; import com.iking.vmware.vim25.mo.PerformanceManager; /** * @description 监控统计vcenter所有对象性能数据 * @date 2017年2月9日11:46:35 * @version 1.1 * @author DiWk */ public class PerformanceCounter { private ConnectedVimServiceBase cs = null; public ConnectedVimServiceBase getCs() { return cs; } public void setCs(ConnectedVimServiceBase cs) { this.cs = cs; } /** * @description 根据属性名称、类型、对象、采集间隔获取所有的性能数据 * @date 2017年2月8日14:37:58 * @return PerformanceMap 性能数据map对象 * @version 1.1 * @author DiWk */ public Map<String, PerformanceManage> getPerfData(String nameInfo, List<String> groupInfo, ManagedEntity mo, Integer interval) { Map<String, PerformanceManage> PerformanceMap = null; Date date = new Date(); Date sTime = new Date(date.getTime() - 24 * 60 * 60 * 1000); Calendar calBegin = Calendar.getInstance(); calBegin.setTime(sTime); Calendar calEnd = Calendar.getInstance(); calEnd.setTime(date); try { PerformanceMap = new HashMap<String, PerformanceManage>(); if (mo != null) { PerformanceManager performanceManager = cs.si.getPerformanceManager(); PerfCounterInfo[] cInfo = performanceManager.getPerfCounter(); Map<Integer, PerfCounterInfo> counters = new HashMap<Integer, PerfCounterInfo>(); for (PerfCounterInfo pcInfo : cInfo) { counters.put(new Integer(pcInfo.getKey()), pcInfo); } PerfMetricId[] listpermeid = performanceManager.queryAvailablePerfMetric(mo, null, null, interval); ArrayList<PerfMetricId> mMetrics = new ArrayList<PerfMetricId>(); if (listpermeid != null) { for (int index = 0; index < listpermeid.length; ++index) { if (counters.containsKey(new Integer(listpermeid[index].getCounterId()))) { mMetrics.add(listpermeid[index]); } } } PerfQuerySpec qSpec = new PerfQuerySpec(); qSpec.setEntity(mo.getMOR()); qSpec.setMetricId(listpermeid); qSpec.setStartTime(calBegin); qSpec.setEndTime(calEnd); qSpec.setIntervalId(interval); qSpec.setFormat("normal"); PerfQuerySpec[] arryQuery = { qSpec }; PerfEntityMetricBase[] pValues = performanceManager.queryPerf(arryQuery); if (pValues == null || pValues.length <= 0) { return null; } PerfSampleInfo[] listperfsinfo = ((PerfEntityMetric) pValues[0]).getSampleInfo(); for (int i = 0; i < pValues.length; i++) { PerfMetricSeries[] listpems = ((PerfEntityMetric) pValues[i]).getValue(); for (int vi = 0; vi < listpems.length; ++vi) { PerfCounterInfo pci = (PerfCounterInfo) counters .get(new Integer(listpems[vi].getId().getCounterId())); if (pci != null) { for (String Info : groupInfo) { PerformanceManage performanceManage = new PerformanceManage(); performanceManage.setStartTime(listperfsinfo[0].getTimestamp().getTime()); performanceManage .setEndTime((listperfsinfo[listperfsinfo.length - 1]).getTimestamp().getTime()); if (pci.getNameInfo().getKey().equalsIgnoreCase(nameInfo) && pci.getGroupInfo().getKey().equalsIgnoreCase(Info)) { if (listpems[vi] instanceof PerfMetricIntSeries) { PerfMetricIntSeries val = (PerfMetricIntSeries) listpems[vi]; long[] lislon = val.getValue(); List<Long> asList = new ArrayList<Long>(); for (Long k : lislon) { asList.add(k); } performanceManage.setPerformanceValues(asList); PerformanceMap.put(Info, performanceManage); } } } } } } } } catch (Exception e) { e.printStackTrace(); } return PerformanceMap; } /** main测试方法 */ public static void main(String[] args) throws Exception { PerformanceCounter performanceCounter2 = new PerformanceCounter(); ConnectedVimServiceBase cs = new ConnectedVimServiceBase(); cs.connect("192.168.1.253", "administrator@vsphere.local", "Iking!@#456"); performanceCounter2.setCs(cs); Folder rootFolder = cs.si.getRootFolder(); HostSystem dataCenter = (HostSystem) new InventoryNavigator(rootFolder) .searchManagedEntity(VsphereConst.HOSTSYSTEM, "192.168.1.254"); List<String> listNm = new ArrayList<String>(); listNm.add("datastore"); Map<String, PerformanceManage> perfData = performanceCounter2.getPerfData("write", listNm, dataCenter, 20); PerformanceManage performanceManage = perfData.get("datastore"); List<Long> performanceValues = performanceManage.getPerformanceValues(); System.out.println(performanceValues.toString()); } }
上面就是通过Java代码实现对vcenter性能的实时监控,相信大家结合官方文档和我分享的电子书一定能实现自己想要的效果。
关于数据存储并没有历史信息的监控,只有实时信息的监控。所以当我统计数据存储的历史信息时,我是累加了所有的虚拟机、主机关于数据存储的数据。我知道这不是一个很好的方法,但是目前没想到更好的。希望和网友分享学习。