Scenario Map: Hyper-V 常见问题排查导航

Product/Service: Windows Server Hyper-V
Scope: Hyper-V 虚拟机管理和运行中的常见故障场景全覆盖
Last Updated: 2026-03-17

1. 场景全景图 (Scenario Overview)

mindmap
  root((Hyper-V<br/>Troubleshooting))
    VM Lifecycle
      A1: VM Won't Start
      A2: VM Stuck in Saved/Paused State
      A3: VMMS/VMWP Crash
    Performance
      B1: VM CPU Bottleneck
      B2: VM Memory Pressure
      B3: VM Storage IO Slow
      B4: VM Network Slow
    Migration
      C1: Live Migration Failure
      C2: Live Migration Slow
      C3: Storage Migration Failure
    Storage
      D1: Checkpoint AVHDX Issues
      D2: VHDX Corruption or Expand Failure
      D3: Pass-through Disk Problems
    Networking
      E1: VM No Network Connectivity
      E2: VM Network Intermittent
      E3: vSwitch Configuration Issues
    Replication
      F1: Replica Initial Replication Failure
      F2: Replica Delta Replication Lag
      F3: Failover Issues

这张全景图覆盖了 Hyper-V 支持中 6 大类 18 个常见场景。根据你遇到的问题症状，在下表中快速定位。

2. 场景识别指南 (How to Identify Your Scenario)

你看到的症状 可能的场景 跳转
VM 启动报错、停在 Starting 状态 VM Won’t Start → 场景 A1
VM 卡在 Saved / Paused / Off (Critical) VM Stuck State → 场景 A2
vmms.exe / vmwp.exe 崩溃、所有 VM 受影响 VMMS/VMWP Crash → 场景 A3
VM 内 CPU 高但 Host 不高 / vCPU 等待时间长 CPU Bottleneck → 场景 B1
VM 内存不足、频繁 Page Fault、Smart Paging Memory Pressure → 场景 B2
VM 磁盘延迟高、IO 慢 Storage IO Slow → 场景 B3
VM 网络吞吐低、丢包、高延迟 Network Slow → 场景 B4
Live Migration 报错回退 LM Failure → 场景 C1
Live Migration 完成但耗时过长 LM Slow → 场景 C2
Checkpoint 删除不了 / AVHDX 不合并 / 链过长 AVHDX Issues → 场景 D1
VHDX 损坏 / 在线扩展失败 VHDX Corruption → 场景 D2
VM 完全没有网络 No Connectivity → 场景 E1
Replica 初始复制失败 Replica Init Fail → 场景 F1
Replica 复制延迟大 Replica Lag → 场景 F2

3. 各场景排查详情 (Scenario Details)

场景 A1: VM 无法启动 (VM Won’t Start)

典型症状： Hyper-V Manager 中点击 Start 后 VM 状态卡在 “Starting” 或直接报错 “Failed to start”

排查逻辑：

先确认 Hypervisor 本身是否正常运行，再检查 VM 的存储和配置文件是否完整可访问，最后排查资源不足和权限问题。

排查流程图：

flowchart TD
    Start["Symptom: VM fails to start"] --> HV{"Hypervisor running?<br/>systeminfo | findstr Hyper-V"}
    HV -->|"Not running"| FixHV["Check BIOS virtualization<br/>bcdedit /set hypervisorlaunchtype auto<br/>Restart host"]
    HV -->|"Running"| VHD{"VHD/VHDX files<br/>accessible?"}
    VHD -->|"Missing/Locked"| FixVHD["Restore files<br/>Check file locks<br/>Verify NTFS permissions"]
    VHD -->|"OK"| Chain{"AVHDX chain<br/>intact?<br/>Get-VHD -Path"}
    Chain -->|"Broken ParentPath"| FixChain["Repair chain:<br/>Set-VHD or manual edit<br/>Merge orphaned AVHDX"]
    Chain -->|"OK"| Mem{"Sufficient host<br/>memory for<br/>Startup Memory?"}
    Mem -->|"Insufficient"| FixMem["Free memory:<br/>Stop other VMs or<br/>reduce Startup Memory"]
    Mem -->|"OK"| Config{"Check VM config<br/>& Event Log<br/>Hyper-V-VMMS/Admin"}
    Config -->|"Error found"| FixConfig["Fix per event details:<br/>NIC mismatch, vSwitch missing,<br/>security policy, etc."]
    Config -->|"No clear error"| Escalate["Collect VMMS + Worker<br/>event logs, escalate"]

关键诊断命令：

检查什么 命令 看什么
Hypervisor 状态 systeminfo \| findstr "Hyper-V" 应显示 “A hypervisor has been detected”
VM 磁盘链 Get-VHD -Path "C:\VMs\disk.avhdx" ParentPath 是否指向存在的文件
VM 状态 Get-VM -Name "TestVM" \| FL * State, Status, MemoryAssigned
事件日志 Event Viewer → Hyper-V-VMMS/Admin 启动失败的具体错误代码

💡 Tips：

  • Tip 1： 如果 VM 有 Checkpoint，AVHDX 链断裂是最常见的启动失败原因。用 Get-VHD 逐个检查 ParentPath
  • Tip 2： 第三方备份软件（如 Veeam、Arcserve）有时会锁定 VHD 文件导致 VM 无法启动
  • ⚠️ 常见误区： 不要直接删除 AVHDX 文件！这会永久丢失 Checkpoint 以来的所有数据

解决方案摘要：

根因 修复方法 验证方式
Hypervisor 未启动 bcdedit /set hypervisorlaunchtype auto + 重启 systeminfo 确认
VHD 文件丢失 从备份恢复 / 修复存储路径 Test-Path 确认文件存在
AVHDX 链断裂 Set-VHD -ParentPath 修复 Get-VHD 验证完整链
Host 内存不足 释放内存或调低 Startup Memory Get-VM 确认启动成功

场景 A2: VM 卡在 Saved/Paused 状态

典型症状： VM 状态显示 Saved、Paused 或 Off (Critical)，无法正常恢复

排查逻辑：

Saved 状态通常因 Host 内存不足或磁盘空间不足导致的自动保存；Paused 状态可能是存储路径不可达；Critical 通常表示 vmwp.exe 已崩溃。

排查流程图：

flowchart TD
    Start["VM in Saved/Paused/Critical state"] --> State{"What state?"}
    State -->|"Saved"| SavedCheck{"Host memory<br/>available?"}
    SavedCheck -->|"Low"| FreeMem["Free memory, then<br/>Start VM normally"]
    SavedCheck -->|"OK"| DelSaved["Delete saved state:<br/>Remove-VMSavedState<br/>then Start VM"]
    State -->|"Paused - Critical"| Storage{"Storage path<br/>accessible?"}
    Storage -->|"Inaccessible"| FixStorage["Reconnect storage<br/>Fix CSV/SAN issues"]
    Storage -->|"OK"| DiskSpace{"Disk space<br/>sufficient?"}
    DiskSpace -->|"Full"| FreeSpace["Expand volume or<br/>clean up disk space"]
    DiskSpace -->|"OK"| Worker["Check vmwp.exe<br/>for this VM"]
    State -->|"Off - Critical"| EventLog["Check Hyper-V-Worker<br/>event log for crash"]
    EventLog --> FixCrash["Restart VM<br/>If recurring, collect dumps"]

💡 Tips：

  • Tip 1： Remove-VMSavedState -VMName "VM1" 可以丢弃保存的状态并让 VM 冷启动
  • Tip 2： VM 自动暂停通常是因为存储路径丢失或磁盘空间不足，检查 CSV 或 SAN 连接状态

场景 A3: VMMS / VMWP 进程崩溃

典型症状： vmms.exe 崩溃影响所有 VM 管理；或单个 vmwp.exe 崩溃导致对应 VM 进入 Critical 状态

排查流程图：

flowchart TD
    Start["VMMS or VMWP crash"] --> Which{"Which process?"}
    Which -->|"vmms.exe"| VMMSImpact["All VMs affected<br/>Cannot manage any VM"]
    VMMSImpact --> RestartSvc["Restart-Service vmms<br/>Check if persists"]
    RestartSvc --> WER["Check WER reports:<br/>C:\\ProgramData\\Microsoft<br/>\\Windows\\WER"]
    Which -->|"vmwp.exe"| VMWPImpact["Single VM in Critical state"]
    VMWPImpact --> IdentifyVM["Identify VM by GUID<br/>Get-VM | FL Name,Id"]
    IdentifyVM --> ThirdParty{"3rd party software<br/>running on host?"}
    ThirdParty -->|"Yes (AV/Backup)"| Exclude["Add exclusions for<br/>VM files and Hyper-V<br/>processes"]
    ThirdParty -->|"No"| Updates{"Recent Windows<br/>Update installed?"}
    Updates -->|"Yes"| KBCheck["Check known issues<br/>for that KB"]
    Updates -->|"No"| Dump["Collect full crash<br/>dump for analysis"]

💡 Tips：

  • Tip 1： 反病毒软件扫描 VHD/VHDX/AVHDX 文件是导致 vmwp.exe 崩溃的常见原因。必须添加排除项
  • ⚠️ 常见误区： 重启 vmms 服务不会影响正在运行的 VM（VM 由各自的 vmwp.exe 进程管理）

场景 B1: VM CPU 瓶颈

典型症状： VM 内应用响应慢，VM 内看 CPU 高，但 Host 物理 CPU 不一定高

排查流程图：

flowchart TD
    Start["VM CPU bottleneck suspected"] --> HostCPU{"Host Physical CPU<br/>utilization > 80%?"}
    HostCPU -->|"Yes"| Overcommit["CPU overcommit:<br/>Too many vCPUs across VMs<br/>Reduce vCPU count or<br/>migrate VMs"]
    HostCPU -->|"No"| WaitTime{"vCPU Wait Time<br/>Per Dispatch<br/>> 50000 ns?"}
    WaitTime -->|"Yes"| ReduceVCPU["CPU scheduling contention:<br/>Reduce total vCPU count<br/>Check NUMA alignment"]
    WaitTime -->|"No"| GuestInternal{"Check inside VM:<br/>Is it a guest OS/<br/>application issue?"}
    GuestInternal --> AppFix["Profile app CPU usage<br/>Check for driver issues<br/>Update Integration Services"]

关键诊断命令：

检查什么 命令/计数器 判断标准
Host CPU \Processor(_Total)\% Processor Time > 80% 持续表示 Host 过载
vCPU 等待 \Hyper-V Hypervisor Virtual Processor(*)\CPU Wait Time Per Dispatch > 50,000 ns = 严重超分
NUMA 对齐 Get-VM \| FL ProcessorCount, NumaAligned 高性能 VM 应为 True

💡 Tips：

  • Tip 1： vCPU 数量不是”越多越好”。过多的 vCPU 会导致 Hypervisor 调度开销增加，反而更慢
  • Tip 2： 使用 Get-VMProcessor 检查 VM 是否启用了 “Processor Compatibility Mode”（会降低 CPU 功能集）

场景 B2: VM 内存压力

典型症状： VM 内存不足告警、应用 OOM、Guest 频繁 Page Fault

排查流程图：

flowchart TD
    Start["VM memory pressure"] --> DynMem{"Dynamic Memory<br/>enabled?"}
    DynMem -->|"Yes"| HostMem{"Host available<br/>memory sufficient?"}
    HostMem -->|"Low < 2GB"| HostPressure["Host memory exhausted<br/>Balance VMs or add RAM"]
    HostMem -->|"OK"| MaxReached{"VM at Maximum<br/>Memory limit?"}
    MaxReached -->|"Yes"| IncreaseMax["Increase Maximum Memory<br/>setting for this VM"]
    MaxReached -->|"No"| Balancer["Check Memory Balancer<br/>Memory Buffer % setting<br/>May need higher priority"]
    DynMem -->|"No - Static"| StaticSize{"Allocated memory<br/>sufficient for workload?"}
    StaticSize -->|"No"| Increase["Increase static memory<br/>Requires VM shutdown"]
    StaticSize -->|"Yes"| GuestLeak["Check guest for<br/>memory leak in<br/>applications/services"]

💡 Tips：

  • Tip 1： Smart Paging 如果频繁触发（磁盘上看到临时 .bin 文件），说明 Host 内存严重不足
  • Tip 2： 对 SQL Server、Exchange 等内存敏感应用，建议使用静态内存而非 Dynamic Memory

场景 B3: VM 存储 IO 缓慢

典型症状： VM 内磁盘延迟高、应用读写慢、Event Log 中有 storvsp 或 disk timeout 事件

排查流程图：

flowchart TD
    Start["VM storage IO slow"] --> AVHDXChain{"AVHDX checkpoint<br/>chain length?"}
    AVHDXChain -->|"> 3 levels"| MergeCP["Merge checkpoints:<br/>Remove-VMCheckpoint<br/>Wait for merge"]
    AVHDXChain -->|"OK"| DiskType{"VHD type?"}
    DiskType -->|"Dynamic expanding"| Convert["Convert to Fixed:<br/>Convert-VHD -VHDType Fixed"]
    DiskType -->|"Fixed"| HostDisk{"Host physical disk<br/>latency high?"}
    HostDisk -->|"Yes"| PhysDisk["Check physical storage:<br/>SAN latency, disk health<br/>RAID rebuild, etc."]
    HostDisk -->|"OK"| AV{"Antivirus scanning<br/>VHD/VHDX files?"}
    AV -->|"Yes"| Exclude["Add AV exclusions<br/>for .vhd/.vhdx/.avhdx"]
    AV -->|"No"| Controller["Check virtual controller:<br/>IDE vs SCSI<br/>Use SCSI for best perf"]

关键诊断命令：

检查什么 命令/计数器 判断标准
AVHDX 链深度 Get-VMCheckpoint -VMName "VM1" > 3 层需要关注
Host 磁盘延迟 \PhysicalDisk(*)\Avg. Disk sec/Read > 20ms 需排查物理存储
VM 虚拟磁盘 IO \Hyper-V Virtual Storage Device(*)\Read/Write Latency 对比基线

💡 Tips：

  • Tip 1： 长期不合并的 Checkpoint 是 VM 存储性能杀手！定期检查并清理不需要的 Checkpoint
  • ⚠️ 常见误区： IDE 控制器在 Gen1 VM 中比 SCSI 慢得多，但只有 IDE 可以做启动盘（Gen1 限制）

场景 B4: VM 网络性能差

典型症状： VM 网络吞吐低、延迟高、丢包

排查流程图：

flowchart TD
    Start["VM network slow/drops"] --> ISCheck{"Integration Services<br/>installed and current?"}
    ISCheck -->|"No/Outdated"| UpdateIS["Update Integration Services<br/>Synthetic NIC requires IS"]
    ISCheck -->|"Yes"| NICType{"NIC type:<br/>Synthetic or Legacy?"}
    NICType -->|"Legacy"| SwitchSyn["Switch to Synthetic NIC<br/>Legacy NIC = emulated = slow"]
    NICType -->|"Synthetic"| VMQ{"VMQ enabled<br/>on host NIC?"}
    VMQ -->|"Disabled"| EnableVMQ["Enable VMQ:<br/>Set-NetAdapterVmq -Enabled $true"]
    VMQ -->|"Enabled"| SRIOV{"SR-IOV available<br/>and enabled?"}
    SRIOV -->|"Not enabled"| CheckHW["If NIC supports SR-IOV,<br/>enable in VM settings"]
    SRIOV -->|"Enabled/N/A"| vRSS{"vRSS enabled<br/>inside VM?"}
    vRSS -->|"No"| EnablevRSS["Enable-NetAdapterRss<br/>inside VM"]
    vRSS -->|"Yes"| PhysNIC["Check physical NIC<br/>errors, driver, firmware"]

💡 Tips：

  • Tip 1： 如果 VM 使用 Legacy NIC（模拟的 DEC 21140 网卡），最高只能到约 100Mbps。必须换 Synthetic NIC
  • Tip 2： SET (Switch Embedded Teaming) 环境中，确保所有网卡型号、驱动版本一致

场景 C1: Live Migration 失败

典型症状： Live Migration 过程中报错回退，VM 仍在源主机运行

排查流程图：

flowchart TD
    Start["Live Migration failed"] --> EventLog["Check Hyper-V-VMMS<br/>event log for error"]
    EventLog --> ErrorType{"Error type?"}
    ErrorType -->|"CPU incompatible"| CPUCompat["Enable Processor<br/>Compatibility Mode<br/>on VM settings"]
    ErrorType -->|"Authentication failed"| Auth{"Auth method?"}
    Auth -->|"CredSSP"| CredSSP["Ensure logged into<br/>source host first<br/>Check CredSSP config"]
    Auth -->|"Kerberos"| Kerberos["Configure Constrained<br/>Delegation in AD<br/>for both host accounts"]
    ErrorType -->|"Network error"| Network["Verify LM network:<br/>Test-NetConnection<br/>Port 6600 (default)"]
    ErrorType -->|"Storage not found"| StorageCheck["Ensure both hosts<br/>can access VM storage<br/>path/CSV/SMB share"]
    ErrorType -->|"Memory insufficient"| DestMem["Free memory on<br/>destination host"]
    ErrorType -->|"Other"| Collect["Collect detailed<br/>event logs from<br/>both hosts"]

关键诊断命令：

检查什么 命令 看什么
兼容性测试 Compare-VM -Name "VM1" -DestinationHost "HV02" 列出所有不兼容项
LM 配置 Get-VMHost \| Select *Migration* 并发数、网络、认证方式
网络连通 Test-NetConnection HV02 -Port 6600 确认 LM 端口可达
委派配置 AD 中检查 Computer Account 的 Delegation 标签 Kerberos Constrained Delegation

💡 Tips：

  • Tip 1： CPU 兼容性问题在异构集群（不同代 CPU 的节点）中最常见。Processor Compatibility Mode 会限制 CPU 功能集到最低公共集
  • Tip 2： Kerberos 认证需要在 AD 中为两台 Host 的计算机账户都配置约束委派
  • ⚠️ 常见误区： CredSSP 认证不支持”第三方发起”的迁移（即不能从管理机远程触发，必须在源主机上操作）

场景 C2: Live Migration 缓慢

典型症状： Migration 不报错，但完成时间远超预期

排查流程图：

flowchart TD
    Start["Live Migration very slow"] --> BW{"LM network<br/>bandwidth?"}
    BW -->|"< 10 Gbps"| UpgradeBW["Use faster network<br/>or enable Compression"]
    BW -->|">= 10 Gbps"| DirtyRate{"VM memory<br/>dirty rate high?<br/>DB/heavy write workload?"}
    DirtyRate -->|"Yes"| Options["Options:<br/>1. Use RDMA/SMB Direct<br/>2. Schedule during low activity<br/>3. Use Quick Migration instead"]
    DirtyRate -->|"No"| Concurrent{"Too many concurrent<br/>migrations?"}
    Concurrent -->|"Yes"| ReduceConc["Reduce concurrent<br/>migration count"]
    Concurrent -->|"No"| RDMA{"RDMA/SMB Direct<br/>available?"}
    RDMA -->|"Yes"| EnableRDMA["Configure SMB Direct<br/>for LM performance option"]
    RDMA -->|"No"| Compress["Enable Compression<br/>performance option"]

💡 Tips：

  • Tip 1： 高内存写入率（Dirty Rate）的 VM（如 OLTP 数据库）可能永远无法达到 Convergence。此时选择 Quick Migration 更实际
  • Tip 2： SMB Direct (RDMA) 是最佳的 LM 传输方式，但需要 RDMA 网卡（如 Mellanox ConnectX）

场景 D1: Checkpoint / AVHDX 问题

典型症状： Checkpoint 删除后 AVHDX 文件未合并、链过长导致性能下降、VM 无法从 Checkpoint 恢复

排查流程图：

flowchart TD
    Start["Checkpoint/AVHDX issue"] --> Symptom{"Symptom?"}
    Symptom -->|"AVHDX not merging"| MergeCheck{"VM running?"}
    MergeCheck -->|"Yes"| LiveMerge["Live Merge in progress?<br/>Check disk IO activity<br/>Wait or shutdown VM<br/>to complete merge"]
    MergeCheck -->|"No"| ManualMerge["Use Edit Disk wizard<br/>Merge child to parent<br/>Start from most recent"]
    Symptom -->|"Chain too long > 3"| Performance["Performance degraded<br/>Plan maintenance window<br/>Delete checkpoints in order<br/>Monitor merge progress"]
    Symptom -->|"Can't delete checkpoint"| Stuck["1. Stop VM if possible<br/>2. Remove-VMCheckpoint<br/>3. If stuck, restart VMMS<br/>4. Check for backup<br/>software locks"]
    Symptom -->|"Orphaned AVHDX files"| Orphaned["Inspect disk chain:<br/>Get-VHD for each .avhdx<br/>Manually merge using<br/>Edit Virtual Hard Disk"]

💡 Tips：

  • Tip 1： 合并顺序必须从最新的（最末端的）AVHDX 开始，逐级向上合并
  • Tip 2： Live Merge 在 VM 运行时可能非常慢（取决于 IO 负载），大型 Checkpoint 建议在维护窗口关机合并
  • ⚠️ 常见误区： 在文件系统层面直接删除 .avhdx 文件 = 永久数据丢失！始终通过 Hyper-V Manager 或 PowerShell 操作

场景 D2: VHDX 损坏或扩展失败

典型症状： VM 内磁盘报 IO 错误、VHDX 在线扩展失败、Compact 操作失败

排查流程图：

flowchart TD
    Start["VHDX corruption or expand failure"] --> Type{"Issue type?"}
    Type -->|"Online expand failed"| ExpandCheck{"Gen1 or Gen2 VM?"}
    ExpandCheck -->|"Gen1 - IDE"| NoExpand["IDE controller does NOT<br/>support online resize<br/>Must shutdown VM first"]
    ExpandCheck -->|"Gen2 - SCSI"| Partition["Expand VHDX succeeded<br/>but OS partition unchanged?<br/>Use Disk Management inside VM<br/>to extend the volume"]
    Type -->|"VHDX corrupt"| Repair["1. Try Hyper-V Inspect Disk<br/>2. Try repair with Edit Disk<br/>3. Restore from backup<br/>if repair fails"]
    Type -->|"Compact failed"| CompactReq["Compact requires:<br/>1. VM powered off<br/>2. All checkpoints merged<br/>3. Sufficient temp disk space"]

💡 Tips：

  • Tip 1： VHDX 在线扩展只改变虚拟磁盘大小，不会自动扩展 Guest 内部的分区。需要在 VM 内使用 Disk Management 扩展
  • ⚠️ 常见误区： Dynamic VHDX 永远不会自动缩小，即使 Guest 删除了大量文件。需要手动 Compact

场景 E1: VM 完全无网络

典型症状： VM 无法 ping 任何地址、网络完全不通

排查流程图：

flowchart TD
    Start["VM has no network"] --> NICCheck{"VM NIC connected<br/>to vSwitch?"}
    NICCheck -->|"No"| ConnectNIC["Connect NIC to<br/>correct vSwitch"]
    NICCheck -->|"Yes"| vSwitchCheck{"vSwitch exists<br/>and functional?"}
    vSwitchCheck -->|"Missing/Error"| RecreateSwitch["Recreate vSwitch<br/>Check physical NIC binding"]
    vSwitchCheck -->|"OK"| ISCheck{"Integration Services<br/>installed?"}
    ISCheck -->|"No"| InstallIS["Install/Update<br/>Integration Services"]
    ISCheck -->|"Yes"| VLAN{"VLAN configured<br/>correctly?"}
    VLAN -->|"Mismatch"| FixVLAN["Correct VLAN ID<br/>on VM NIC settings"]
    VLAN -->|"OK"| GuestIP{"Guest IP config<br/>correct?"}
    GuestIP -->|"No"| FixIP["Fix IP/DHCP config<br/>inside VM"]
    GuestIP -->|"OK"| Firewall["Check guest firewall<br/>and host-level<br/>port ACLs"]

💡 Tips：

  • Tip 1： 创建 External vSwitch 时绑定物理 NIC，会中断该 NIC 的管理连接。务必保留独立管理 NIC 或使用 SET
  • Tip 2： 如果 vSwitch 绑定的物理 NIC 驱动更新失败，vSwitch 会断开所有 VM 的网络

场景 F1: Replica 初始复制失败

典型症状： 配置 Hyper-V Replica 后初始复制报错、连接不上、认证失败

排查流程图：

flowchart TD
    Start["Replica initial replication failed"] --> AuthType{"Auth type?"}
    AuthType -->|"Kerberos/HTTP"| Port80{"Port 80 open<br/>between hosts?"}
    Port80 -->|"Blocked"| OpenFW["Open firewall rule:<br/>Hyper-V Replica HTTP"]
    Port80 -->|"Open"| ReplicaEnabled{"Replica enabled<br/>on target host?"}
    AuthType -->|"Certificate/HTTPS"| Port443{"Port 443 open?"}
    Port443 -->|"Blocked"| OpenFW443["Open firewall rule:<br/>Hyper-V Replica HTTPS"]
    Port443 -->|"Open"| CertValid{"Certificate valid<br/>and trusted?"}
    CertValid -->|"No"| FixCert["Install/renew cert<br/>Ensure mutual trust"]
    CertValid -->|"Yes"| ReplicaEnabled
    ReplicaEnabled -->|"No"| EnableReplica["Enable Replication<br/>on target Hyper-V host"]
    ReplicaEnabled -->|"Yes"| DNS{"DNS resolution<br/>working between hosts?"}
    DNS -->|"No"| FixDNS["Fix DNS or use<br/>hosts file entries"]
    DNS -->|"OK"| StorageSpace["Ensure sufficient<br/>storage on target"]

💡 Tips：

  • Tip 1： 大型 VM（TB 级别磁盘）可以选择”初始复制导出到外部媒体”方式，避免网络传输瓶颈
  • Tip 2： 跨域场景必须使用证书认证（HTTPS），Kerberos 只适用于同域环境

场景 F2: Replica Delta 复制延迟

典型症状： Replica 健康状态显示 Warning/Critical，复制延迟大于配置的 RPO

排查流程图：

flowchart TD
    Start["Replica replication lag"] --> BW{"Network bandwidth<br/>sufficient?"}
    BW -->|"Saturated"| UpBW["Upgrade bandwidth<br/>or reduce replication<br/>frequency"]
    BW -->|"OK"| ChangeRate{"VM change rate<br/>too high?"}
    ChangeRate -->|"Yes"| Freq["Increase replication<br/>interval (e.g., 30s to 5min)<br/>or reduce VM IO"]
    ChangeRate -->|"No"| TargetDisk{"Target storage<br/>IO keeping up?"}
    TargetDisk -->|"Slow"| UpgradeTarget["Upgrade target<br/>storage performance"]
    TargetDisk -->|"OK"| Health["Check Replica health:<br/>Measure-VMReplication"]

💡 Tips：

  • Tip 1： Measure-VMReplication -VMName "VM1" 可以查看详细的复制健康状态、延迟、大小
  • ⚠️ 常见误区： 30 秒复制频率听起来很好，但对于高变更率的 VM 可能导致持续的网络/存储压力

4. 通用排查工具箱 (Universal Toolkit)

日志收集

目的 命令/工具 说明
Hyper-V 管理事件 Event Viewer → Hyper-V-VMMS/Admin VM 管理相关错误
VM Worker 事件 Event Viewer → Hyper-V-Worker/Admin 单个 VM 运行时错误
Hypervisor 事件 Event Viewer → Hyper-V-Hypervisor/Admin Hypervisor 底层错误
导出所有 Hyper-V 日志 Get-WinEvent -ListLog *Hyper-V* \| ForEach { Get-WinEvent -LogName $_.LogName -MaxEvents 100 } 批量导出
性能数据 logman create counter "HyperV-Perf" -cf HyperVCounters.txt -si 5 -max 512 自定义计数器集

诊断命令

目的 命令 说明
列出所有 VM 状态 Get-VM \| Format-Table Name, State, CPUUsage, MemoryAssigned -AutoSize 快速概览
检查 VHD 链 Get-VHD -Path "path.vhdx" 查看 ParentPath、大小、格式
检查 LM 兼容性 Compare-VM -Name "VM1" -DestinationHost "HV02" 迁移前检测
Replica 健康 Measure-VMReplication -VMName "VM1" 复制状态和延迟
Host 资源概览 Get-VMHost \| Select *Memory*, *Migration* Host 配置检查
VM 网络详情 Get-VMNetworkAdapter -VMName "VM1" \| FL * 网卡配置详情

常用注册表/配置

配置项 路径/键值 作用
Hypervisor 启动类型 bcdedit /enum {hypervisorlaunchtype} Auto = Hypervisor 随系统启动
VM 配置文件路径 C:\ProgramData\Microsoft\Windows\Hyper-V\ 默认 VM 配置位置
VHD 默认路径 C:\Users\Public\Documents\Hyper-V\Virtual Hard Disks\ 默认 VHD 存储位置
Smart Paging 路径 VM 设置 → Smart Paging File Location 建议放在 SSD 上

5. 跨场景关联 (Cross-Scenario Relationships)

flowchart LR
    A1["A1: VM Won't Start"] -.->|"Root cause may be"| D1["D1: AVHDX Issues"]
    A1 -.->|"Check also"| B2["B2: Memory Pressure"]
    B3["B3: Storage IO Slow"] -.->|"Often caused by"| D1
    C1["C1: LM Failure"] -.->|"May relate to"| B2
    C2["C2: LM Slow"] -.->|"Bottleneck similar to"| B3
    E1["E1: No Network"] -.->|"After"| C1
    A3["A3: VMWP Crash"] -.->|"May cause"| A1
    F2["F2: Replica Lag"] -.->|"Bottleneck at"| B3

关键关联提示：

  • VM 启动失败（A1）最常见根因是 AVHDX 问题（D1）
  • 存储 IO 慢（B3）通常和 Checkpoint 链过长（D1）同时出现
  • Live Migration 后 VM 网络断开（E1）通常是 vSwitch 配置在目标 Host 不一致
  • VMWP 崩溃（A3）会导致 VM 进入 Critical 状态，表现为”VM 无法启动”（A1）

6. 参考资料 (References)

English Version

Scenario Map: Hyper-V Common Troubleshooting Navigation

Product/Service: Windows Server Hyper-V
Scope: Comprehensive coverage of common Hyper-V VM management and runtime fault scenarios
Last Updated: 2026-03-17

1. Scenario Overview

mindmap
  root((Hyper-V<br/>Troubleshooting))
    VM Lifecycle
      A1: VM Won't Start
      A2: VM Stuck in Saved/Paused State
      A3: VMMS/VMWP Crash
    Performance
      B1: VM CPU Bottleneck
      B2: VM Memory Pressure
      B3: VM Storage IO Slow
      B4: VM Network Slow
    Migration
      C1: Live Migration Failure
      C2: Live Migration Slow
      C3: Storage Migration Failure
    Storage
      D1: Checkpoint AVHDX Issues
      D2: VHDX Corruption or Expand Failure
      D3: Pass-through Disk Problems
    Networking
      E1: VM No Network Connectivity
      E2: VM Network Intermittent
      E3: vSwitch Configuration Issues
    Replication
      F1: Replica Initial Replication Failure
      F2: Replica Delta Replication Lag
      F3: Failover Issues

This overview covers 6 categories with 18 common scenarios. Use the symptom table below to quickly locate your issue.

2. How to Identify Your Scenario

Symptom You See Likely Scenario Jump To
VM start fails, stuck in “Starting” state VM Won’t Start → Scenario A1
VM stuck in Saved / Paused / Off (Critical) VM Stuck State → Scenario A2
vmms.exe / vmwp.exe crashes, all VMs affected VMMS/VMWP Crash → Scenario A3
VM CPU high but Host not / long vCPU wait CPU Bottleneck → Scenario B1
VM out of memory, frequent page faults Memory Pressure → Scenario B2
VM high disk latency, slow IO Storage IO Slow → Scenario B3
VM low throughput, packet loss, high latency Network Slow → Scenario B4
Live Migration errors and rolls back LM Failure → Scenario C1
Live Migration completes but takes too long LM Slow → Scenario C2
Checkpoint won’t delete / AVHDX won’t merge AVHDX Issues → Scenario D1
VHDX corruption / online expand failure VHDX Corruption → Scenario D2
VM has no network at all No Connectivity → Scenario E1
Replica initial replication fails Replica Init Fail → Scenario F1
Replica replication lag high Replica Lag → Scenario F2

3. Scenario Details

Scenario A1: VM Won’t Start

Typical Symptom: Click Start in Hyper-V Manager, VM stays in “Starting” state or shows “Failed to start”

Troubleshooting Logic:

First verify the Hypervisor itself is running, then check VM storage and config file accessibility, finally investigate resource shortages and permission issues.

Troubleshooting Flowchart:

flowchart TD
    Start["Symptom: VM fails to start"] --> HV{"Hypervisor running?<br/>systeminfo | findstr Hyper-V"}
    HV -->|"Not running"| FixHV["Check BIOS virtualization<br/>bcdedit /set hypervisorlaunchtype auto<br/>Restart host"]
    HV -->|"Running"| VHD{"VHD/VHDX files<br/>accessible?"}
    VHD -->|"Missing/Locked"| FixVHD["Restore files<br/>Check file locks<br/>Verify NTFS permissions"]
    VHD -->|"OK"| Chain{"AVHDX chain<br/>intact?<br/>Get-VHD -Path"}
    Chain -->|"Broken ParentPath"| FixChain["Repair chain:<br/>Set-VHD or manual edit<br/>Merge orphaned AVHDX"]
    Chain -->|"OK"| Mem{"Sufficient host<br/>memory for<br/>Startup Memory?"}
    Mem -->|"Insufficient"| FixMem["Free memory:<br/>Stop other VMs or<br/>reduce Startup Memory"]
    Mem -->|"OK"| Config{"Check VM config<br/>and Event Log<br/>Hyper-V-VMMS/Admin"}
    Config -->|"Error found"| FixConfig["Fix per event details:<br/>NIC mismatch, vSwitch missing,<br/>security policy, etc."]
    Config -->|"No clear error"| Escalate["Collect VMMS + Worker<br/>event logs, escalate"]

Key Diagnostic Commands:

What to Check Command What to Look For
Hypervisor status systeminfo \| findstr "Hyper-V" Should show “A hypervisor has been detected”
VM disk chain Get-VHD -Path "C:\VMs\disk.avhdx" ParentPath points to existing file
VM status Get-VM -Name "TestVM" \| FL * State, Status, MemoryAssigned
Event logs Event Viewer → Hyper-V-VMMS/Admin Specific error code for startup failure

💡 Tips:

  • Tip 1: AVHDX chain breaks are the most common cause of startup failure when checkpoints exist. Check ParentPath for each with Get-VHD
  • Tip 2: Third-party backup software (Veeam, Arcserve, etc.) sometimes locks VHD files, preventing VM start
  • ⚠️ Common Mistake: Never delete AVHDX files directly! This permanently loses all data since that checkpoint

Solution Summary:

Root Cause Fix Verification
Hypervisor not started bcdedit /set hypervisorlaunchtype auto + reboot systeminfo confirms
VHD file missing Restore from backup / fix storage path Test-Path confirms file exists
AVHDX chain broken Set-VHD -ParentPath to repair Get-VHD verifies complete chain
Host memory insufficient Free memory or reduce Startup Memory Get-VM confirms successful start

Scenario A2: VM Stuck in Saved/Paused State

Typical Symptom: VM shows Saved, Paused, or Off (Critical) and won’t resume normally

Troubleshooting Flowchart:

flowchart TD
    Start["VM in Saved/Paused/Critical state"] --> State{"What state?"}
    State -->|"Saved"| SavedCheck{"Host memory<br/>available?"}
    SavedCheck -->|"Low"| FreeMem["Free memory, then<br/>Start VM normally"]
    SavedCheck -->|"OK"| DelSaved["Delete saved state:<br/>Remove-VMSavedState<br/>then Start VM"]
    State -->|"Paused - Critical"| Storage{"Storage path<br/>accessible?"}
    Storage -->|"Inaccessible"| FixStorage["Reconnect storage<br/>Fix CSV/SAN issues"]
    Storage -->|"OK"| DiskSpace{"Disk space<br/>sufficient?"}
    DiskSpace -->|"Full"| FreeSpace["Expand volume or<br/>clean up disk space"]
    DiskSpace -->|"OK"| Worker["Check vmwp.exe<br/>for this VM"]
    State -->|"Off - Critical"| EventLog["Check Hyper-V-Worker<br/>event log for crash"]

💡 Tips:

  • Tip 1: Remove-VMSavedState -VMName "VM1" discards saved state and allows cold boot
  • Tip 2: Auto-pause usually means storage path loss or insufficient disk space — check CSV or SAN status

Scenario A3: VMMS / VMWP Process Crash

Typical Symptom: vmms.exe crash affects all VM management; single vmwp.exe crash puts one VM in Critical state

Troubleshooting Flowchart:

flowchart TD
    Start["VMMS or VMWP crash"] --> Which{"Which process?"}
    Which -->|"vmms.exe"| RestartSvc["Restart-Service vmms<br/>Check if crash persists"]
    RestartSvc --> WER["Check WER reports:<br/>C:\\ProgramData\\Microsoft<br/>\\Windows\\WER"]
    Which -->|"vmwp.exe"| IdentifyVM["Identify VM by GUID<br/>Get-VM | FL Name,Id"]
    IdentifyVM --> ThirdParty{"3rd party software<br/>on host?"}
    ThirdParty -->|"Yes"| Exclude["Add AV/backup exclusions<br/>for VM files and processes"]
    ThirdParty -->|"No"| Dump["Collect full crash<br/>dump for analysis"]

💡 Tips:

  • Tip 1: Antivirus scanning VHD/VHDX/AVHDX files is a common cause of vmwp.exe crashes. Must add exclusions
  • ⚠️ Common Mistake: Restarting vmms service does NOT affect running VMs (each VM has its own vmwp.exe)

Scenario B1: VM CPU Bottleneck

Troubleshooting Flowchart:

flowchart TD
    Start["VM CPU bottleneck"] --> HostCPU{"Host CPU > 80%?"}
    HostCPU -->|"Yes"| Overcommit["CPU overcommit<br/>Reduce vCPU count<br/>or migrate VMs"]
    HostCPU -->|"No"| WaitTime{"vCPU Wait Time<br/>> 50000 ns?"}
    WaitTime -->|"Yes"| ReduceVCPU["Scheduling contention<br/>Reduce total vCPUs<br/>Check NUMA alignment"]
    WaitTime -->|"No"| GuestIssue["Guest OS/app issue<br/>Profile CPU usage<br/>Update Integration Services"]

💡 Tips:

  • Tip 1: More vCPUs is NOT always better. Excessive vCPUs increase Hypervisor scheduling overhead
  • Tip 2: Check Get-VMProcessor for “Processor Compatibility Mode” which may reduce CPU features

Scenario B2: VM Memory Pressure

Troubleshooting Flowchart:

flowchart TD
    Start["VM memory pressure"] --> DynMem{"Dynamic Memory?"}
    DynMem -->|"Yes"| HostMem{"Host available<br/>memory > 2GB?"}
    HostMem -->|"Low"| HostFix["Host exhausted<br/>Balance VMs or add RAM"]
    HostMem -->|"OK"| MaxReached{"VM at Max Memory?"}
    MaxReached -->|"Yes"| IncMax["Increase Max Memory"]
    MaxReached -->|"No"| Balancer["Check buffer %<br/>and memory priority"]
    DynMem -->|"No"| StaticCheck{"Static RAM sufficient?"}
    StaticCheck -->|"No"| IncStatic["Increase static memory<br/>Requires VM shutdown"]
    StaticCheck -->|"Yes"| Leak["Check guest for<br/>memory leak"]

💡 Tips:

  • Tip 1: Frequent Smart Paging (temporary .bin files on disk) indicates severe host memory shortage
  • Tip 2: For SQL Server, Exchange — use static memory instead of Dynamic Memory

Scenario B3: VM Storage IO Slow

Troubleshooting Flowchart:

flowchart TD
    Start["VM storage IO slow"] --> Chain{"AVHDX chain > 3?"}
    Chain -->|"Yes"| Merge["Merge checkpoints"]
    Chain -->|"OK"| DiskType{"Dynamic VHD?"}
    DiskType -->|"Yes"| Convert["Convert to Fixed"]
    DiskType -->|"Fixed"| HostDisk{"Host disk<br/>latency high?"}
    HostDisk -->|"Yes"| PhysFix["Check physical storage"]
    HostDisk -->|"OK"| AV{"AV scanning<br/>VHD files?"}
    AV -->|"Yes"| Exclude["Add AV exclusions"]
    AV -->|"No"| Controller["Check IDE vs SCSI<br/>Use SCSI for best perf"]

💡 Tips:

  • Tip 1: Unmerged checkpoints are the #1 killer of VM storage performance
  • ⚠️ Common Mistake: IDE controller in Gen1 VMs is significantly slower than SCSI

Scenario B4: VM Network Performance Issues

Troubleshooting Flowchart:

flowchart TD
    Start["VM network slow"] --> IS{"Integration Services<br/>current?"}
    IS -->|"No"| Update["Update IS"]
    IS -->|"Yes"| NIC{"Synthetic or Legacy NIC?"}
    NIC -->|"Legacy"| Switch["Switch to Synthetic"]
    NIC -->|"Synthetic"| VMQ{"VMQ enabled?"}
    VMQ -->|"No"| EnableVMQ["Enable VMQ"]
    VMQ -->|"Yes"| SRIOV{"SR-IOV available?"}
    SRIOV -->|"No"| vRSS["Enable vRSS inside VM"]
    SRIOV -->|"Yes/N/A"| PhysNIC["Check physical NIC<br/>errors, driver, firmware"]

💡 Tips:

  • Tip 1: Legacy NIC (emulated DEC 21140) caps at ~100 Mbps. Must use Synthetic NIC
  • Tip 2: In SET environments, ensure all NIC models and driver versions match

Scenario C1: Live Migration Failure

Troubleshooting Flowchart:

flowchart TD
    Start["LM failed"] --> Error{"Error type<br/>from Event Log?"}
    Error -->|"CPU incompatible"| CPU["Enable Processor<br/>Compatibility Mode"]
    Error -->|"Auth failed"| Auth["Fix CredSSP or<br/>Kerberos Delegation"]
    Error -->|"Network error"| Net["Test-NetConnection<br/>Port 6600"]
    Error -->|"Storage not found"| Storage["Verify both hosts<br/>can access VM storage"]
    Error -->|"Memory insufficient"| Mem["Free memory on<br/>destination host"]

Key Diagnostic Commands:

What to Check Command What to Look For
Compatibility Compare-VM -Name "VM1" -DestinationHost "HV02" Lists all incompatibilities
LM config Get-VMHost \| Select *Migration* Concurrent count, network, auth
Connectivity Test-NetConnection HV02 -Port 6600 LM port reachable

💡 Tips:

  • Tip 1: CPU compatibility issues are most common in heterogeneous clusters (different CPU generations)
  • ⚠️ Common Mistake: CredSSP doesn’t support “third-party initiated” migration (can’t remotely trigger from management station)

Scenario C2: Live Migration Slow

Troubleshooting Flowchart:

flowchart TD
    Start["LM very slow"] --> BW{"LM network < 10 Gbps?"}
    BW -->|"Yes"| Upgrade["Upgrade network or<br/>enable Compression"]
    BW -->|"No"| Dirty{"High memory<br/>dirty rate?"}
    Dirty -->|"Yes"| Options["Use RDMA/SMB Direct<br/>or Quick Migration"]
    Dirty -->|"No"| Concurrent{"Too many<br/>concurrent migrations?"}
    Concurrent -->|"Yes"| Reduce["Reduce concurrent count"]
    Concurrent -->|"No"| RDMA["Enable SMB Direct<br/>for best performance"]

💡 Tips:

  • Tip 1: VMs with high dirty rates (OLTP databases) may never converge. Use Quick Migration instead
  • Tip 2: SMB Direct (RDMA) is the optimal LM transport, but requires RDMA NICs

Scenario D1: Checkpoint / AVHDX Issues

Troubleshooting Flowchart:

flowchart TD
    Start["Checkpoint/AVHDX issue"] --> Symptom{"Symptom?"}
    Symptom -->|"Not merging"| VM{"VM running?"}
    VM -->|"Yes"| Wait["Live Merge in progress<br/>Wait or shut down VM"]
    VM -->|"No"| Manual["Edit Disk wizard<br/>Merge child to parent"]
    Symptom -->|"Chain too long"| Plan["Plan maintenance<br/>Delete checkpoints in order"]
    Symptom -->|"Can't delete"| Stuck["Stop VM if possible<br/>Remove-VMCheckpoint<br/>Restart VMMS if stuck"]

💡 Tips:

  • Tip 1: Always merge from the newest (outermost) AVHDX first, working up the chain
  • ⚠️ Common Mistake: Directly deleting .avhdx files in file system = permanent data loss

Scenario D2: VHDX Corruption or Expand Failure

Troubleshooting Flowchart:

flowchart TD
    Start["VHDX issue"] --> Type{"Issue type?"}
    Type -->|"Online expand failed"| Gen{"Gen1 IDE or Gen2 SCSI?"}
    Gen -->|"Gen1 IDE"| NoOnline["IDE doesn't support<br/>online resize<br/>Shutdown VM first"]
    Gen -->|"Gen2 SCSI"| Partition["VHDX expanded OK<br/>But extend partition<br/>inside VM with<br/>Disk Management"]
    Type -->|"Corruption"| Repair["Inspect Disk → Edit Disk<br/>Or restore from backup"]
    Type -->|"Compact failed"| Reqs["Requires: VM off,<br/>checkpoints merged,<br/>sufficient temp space"]

💡 Tips:

  • Tip 1: VHDX online expand only changes virtual disk size, does NOT auto-extend the guest partition
  • ⚠️ Common Mistake: Dynamic VHDX never auto-shrinks, even after deleting large files in guest. Manual Compact needed

Scenario E1: VM No Network Connectivity

Troubleshooting Flowchart:

flowchart TD
    Start["VM no network"] --> NIC{"VM NIC connected<br/>to vSwitch?"}
    NIC -->|"No"| Connect["Connect to correct vSwitch"]
    NIC -->|"Yes"| Switch{"vSwitch functional?"}
    Switch -->|"Error"| Recreate["Recreate vSwitch"]
    Switch -->|"OK"| IS{"Integration Services?"}
    IS -->|"No"| Install["Install/Update IS"]
    IS -->|"Yes"| VLAN{"VLAN correct?"}
    VLAN -->|"Mismatch"| Fix["Correct VLAN ID"]
    VLAN -->|"OK"| IP{"Guest IP correct?"}
    IP -->|"No"| FixIP["Fix IP/DHCP config"]
    IP -->|"OK"| FW["Check guest firewall<br/>and port ACLs"]

💡 Tips:

  • Tip 1: Creating an External vSwitch bound to a physical NIC interrupts that NIC’s management connection. Keep a separate management NIC
  • Tip 2: If the physical NIC driver update fails, vSwitch disconnects ALL VMs

Scenario F1: Replica Initial Replication Failure

Troubleshooting Flowchart:

flowchart TD
    Start["Replica init failed"] --> Auth{"Auth type?"}
    Auth -->|"Kerberos/HTTP"| Port80{"Port 80 open?"}
    Port80 -->|"No"| FW["Open Hyper-V Replica<br/>HTTP firewall rule"]
    Port80 -->|"Yes"| Enabled{"Replica enabled<br/>on target?"}
    Auth -->|"Certificate/HTTPS"| Port443{"Port 443 open?"}
    Port443 -->|"No"| FW443["Open HTTPS rule"]
    Port443 -->|"Yes"| Cert{"Cert valid/trusted?"}
    Cert -->|"No"| FixCert["Install/renew cert"]
    Cert -->|"Yes"| Enabled
    Enabled -->|"No"| Enable["Enable on target"]
    Enabled -->|"Yes"| DNS["Check DNS resolution<br/>and storage space"]

💡 Tips:

  • Tip 1: For large VMs (TB-scale), use “Export to External Media” for initial replication to avoid network bottleneck
  • Tip 2: Cross-domain scenarios must use certificate auth (HTTPS); Kerberos only works within the same domain

Scenario F2: Replica Delta Replication Lag

Troubleshooting Flowchart:

flowchart TD
    Start["Replica lag"] --> BW{"Network saturated?"}
    BW -->|"Yes"| Fix["Upgrade bandwidth or<br/>reduce replication frequency"]
    BW -->|"No"| Change{"VM change rate<br/>too high?"}
    Change -->|"Yes"| Adjust["Increase replication<br/>interval or reduce VM IO"]
    Change -->|"No"| Target{"Target storage slow?"}
    Target -->|"Yes"| Upgrade["Upgrade target storage"]
    Target -->|"No"| Health["Measure-VMReplication<br/>for detailed health"]

💡 Tips:

  • Tip 1: Measure-VMReplication -VMName "VM1" shows detailed replication health, lag, and size
  • ⚠️ Common Mistake: 30-second frequency sounds great but can cause sustained network/storage pressure for high-change-rate VMs

4. Universal Toolkit

Log Collection

Purpose Command/Tool Description
Hyper-V management events Event Viewer → Hyper-V-VMMS/Admin VM management errors
VM Worker events Event Viewer → Hyper-V-Worker/Admin Individual VM runtime errors
Hypervisor events Event Viewer → Hyper-V-Hypervisor/Admin Low-level Hypervisor errors
Export all Hyper-V logs Get-WinEvent -ListLog *Hyper-V* \| ForEach { Get-WinEvent -LogName $_.LogName -MaxEvents 100 } Batch export

Diagnostic Commands

Purpose Command Description
List all VM states Get-VM \| FT Name, State, CPUUsage, MemoryAssigned Quick overview
Check VHD chain Get-VHD -Path "path.vhdx" View ParentPath, size, format
LM compatibility Compare-VM -Name "VM1" -DestinationHost "HV02" Pre-migration check
Replica health Measure-VMReplication -VMName "VM1" Replication status and lag
Host overview Get-VMHost \| Select *Memory*, *Migration* Host config check

5. Cross-Scenario Relationships

flowchart LR
    A1["A1: VM Won't Start"] -.->|"Root cause may be"| D1["D1: AVHDX Issues"]
    A1 -.->|"Check also"| B2["B2: Memory Pressure"]
    B3["B3: Storage IO Slow"] -.->|"Often caused by"| D1
    C1["C1: LM Failure"] -.->|"May relate to"| B2
    C2["C2: LM Slow"] -.->|"Bottleneck similar to"| B3
    E1["E1: No Network"] -.->|"After"| C1
    A3["A3: VMWP Crash"] -.->|"May cause"| A1
    F2["F2: Replica Lag"] -.->|"Bottleneck at"| B3

Key relationships:

  • VM startup failure (A1) most commonly caused by AVHDX issues (D1)
  • Storage IO slow (B3) often coincides with long checkpoint chains (D1)
  • Post-Live Migration network loss (E1) usually due to vSwitch config mismatch on target host
  • VMWP crash (A3) puts VM in Critical state, manifesting as “VM won’t start” (A1)

6. References