为了模拟云环境，我们使用Cloudsim构建了一个包含10个主机和30个虚拟机的云环境。其中，每个主机包含4个CPU、8GB内存和100GB存储空间；每个虚拟机包含1个CPU、2GB内存和20GB存储空间。 2.运行实验 在不同负载情况下，我们分别测试了基于多维QoS的虚拟机调度算法和传统的虚拟机调度算法，并记录了以下指标： VM启动时间 VM迁移时间 网络带宽 资源利用率 响应时间 具体而言，我们分别模拟了以下负载情况： 低负载：每个主机上运行3个虚拟机。 中等负载：每个主机上运行6个虚拟机。 高负载：每个主机上运行9个虚拟机。 请书写以上四个代码，并且每个代码用回车隔开。

首先，我们需要定义主机和虚拟机的参数，后创建主机和虚拟机对象。以下是创建主机和虚拟机的代码：

int numHosts = 10;
int numVMs = 30;

List<Host> hostList = new ArrayList<Host>();
List<Vm> vmList = new ArrayList<Vm>();

for (int i = 0; i < numHosts; i++) {
    List<Pe> peList = new ArrayList<Pe>();
    for (int j = 0; j < 4; j++) {
        peList.add(new Pe(j, new PeProvisionerSimple(1000))); // 4个CPU，每个CPU 1000 MIPS
    }
    int ram = 8192; // 8GB内存
    long storage = 100000; // 100GB存储空间
    int bw = 10000; // 带宽

    hostList.add(new Host(i, new RamProvisionerSimple(ram), new BwProvisionerSimple(bw), storage, peList, new VmSchedulerSpaceShared(peList)));
}

for (int i = 0; i < numVMs; i++) {
    int mips = 1000; // 1个CPU，每个CPU 1000 MIPS
    int ram = 2048; // 2GB内存
    long storage = 20000; // 20GB存储空间
    int bw = 1000; // 带宽

    Vm vm = new Vm(i, brokerId, mips, 1, ram, bw, storage, "Xen", new CloudletSchedulerTimeShared());
    vmList.add(vm);
}

接下来，我们可以根据不同的负载情况进行实验，并记录指标。以下是低、中等和高负载情况下的代码示例：

1. 低负载：

// 设置每个主机上运行的虚拟机数量
int vmCountLow = 3;

for (int i = 0; i < numHosts; i++) {
    for (int j = 0; j < vmCountLow; j++) {
        broker.bindVmToHost(vmList.get(j), hostList.get(i));
        // 设置虚拟机的任务，执行一段时间
        cloudletList.get(j).setVmId(vmList.get(j).getId());
        broker.submitCloudletList(cloudletList.subList(j, j+1));
    }
}

CloudSim.startSimulation();
// 等待仿真结束

// 记录指标
double vmStartupTimeLow = calculateVmStartupTime(vmList.subList(0, vmCountLow));
double vmMigrationTimeLow = calculateVmMigrationTime(vmList.subList(0, vmCountLow));
double networkBandwidthLow = calculateNetworkBandwidth(vmList.subList(0, vmCountLow));
double resourceUtilizationLow = calculateResourceUtilization(hostList);
double responseTimeLow = calculateResponseTime(cloudletList.subList(0, vmCountLow));

// 打印指标
System.out.println("低负载情况下的指标：");
System.out.println("VM启动时间：" + vmStartupTimeLow);
System.out.println("VM迁移时间：" + vmMigrationTimeLow);
System.out.println("网络带宽：" + networkBandwidthLow);
System.out.println("资源利用率：" + resourceUtilizationLow);
System.out.println("响应时间：" + responseTimeLow);

2. 中等负载：

// 设置每个主机上运行的虚拟机数量
int vmCountMedium = 6;

for (int i = 0; i < numHosts; i++) {
    for (int j = 0; j < vmCountMedium; j++) {
        broker.bindVmToHost(vmList.get(j), hostList.get(i));
        // 设置虚拟机的任务，执行一段时间
        cloudletList.get(j).setVmId(vmList.get(j).getId());
        broker.submitCloudletList(cloudletList.subList(j, j+1));
    }
}

CloudSim.startSimulation();
// 等待仿真结束

// 记录指标
double vmStartupTimeMedium = calculateVmStartupTime(vmList.subList(0, vmCountMedium));
double vmMigrationTimeMedium = calculateVmMigrationTime(vmList.subList(0, vmCountMedium));
double networkBandwidthMedium = calculateNetworkBandwidth(vmList.subList(0, vmCountMedium));
double resourceUtilizationMedium = calculateResourceUtilization(hostList);
double responseTimeMedium = calculateResponseTime(cloudletList.subList(0, vmCountMedium));

// 打印指标
System.out.println("中等负载情况下的指标：");
System.out.println("VM启动时间：" + vmStartupTimeMedium);
System.out.println("VM迁移时间：" + vmMigrationTimeMedium);
System.out.println("网络带宽：" + networkBandwidthMedium);
System.out.println("资源利用率：" + resourceUtilizationMedium);
System.out.println("响应时间：" + responseTimeMedium);

3. 高负载：

// 设置每个主机上运行的虚拟机数量
int vmCountHigh = 9;

for (int i = 0; i < numHosts; i++) {
    for (int j = 0; j < vmCountHigh; j++) {
        broker.bindVmToHost(vmList.get(j), hostList.get(i));
        // 设置虚拟机的任务，执行一段时间
        cloudletList.get(j).setVmId(vmList.get(j).getId());
        broker.submitCloudletList(cloudletList.subList(j, j+1));
    }
}

CloudSim.startSimulation();
// 等待仿真结束

// 记录指标
double vmStartupTimeHigh = calculateVmStartupTime(vmList.subList(0, vmCountHigh));
double vmMigrationTimeHigh = calculateVmMigrationTime(vmList.subList(0, vmCountHigh));
double networkBandwidthHigh = calculateNetworkBandwidth(vmList.subList(0, vmCountHigh));
double resourceUtilizationHigh = calculateResourceUtilization(hostList);
double responseTimeHigh = calculateResponseTime(cloudletList.subList(0, vmCountHigh));

// 打印指标
System.out.println("高负载情况下的指标：");
System.out.println("VM启动时间：" + vmStartupTimeHigh);
System.out.println("VM迁移时间：" + vmMigrationTimeHigh);
System.out.println("网络带宽：" + networkBandwidthHigh);
System.out.println("资源利用率：" + resourceUtilizationHigh);
System.out.println("响应时间：" + responseTimeHigh);

以上是基于多维QoS的虚拟机调度算法和传统的虚拟机调度算法的实验代码示例。你可以根据自己的需求进行修改和扩展。注意，在运行实验之前，你还需要定义和初始化CloudSim的一些参数，如初始化CloudSim和创建Datacenter、DatacenterBroker等。希望对你有所帮助！