大网数据广告活动
Hyper-V与虚拟化

Guidance for troubleshooting high-availability virtual machines

· 52CE 我爱测

Helps you troubleshoot issues that affect high-availability (HA) virtual machines (VMs) in a Windows Server failover cluster environment.

Summary

This article helps you troubleshoot issues that affect high-availability (HA) virtual machines (VMs) in a Windows Server failover cluster environment. Use this guidance if you experience issues such as VMs that don't migrate, cluster resources that become unresponsive, or storage and communication errors that affect VM availability.

The article is organized into the following sections:

  • Troubleshooting checklist: Guidance to help you isolate the issue and fix the most common underlying issues before you begin deeper investigation.
  • Common issues and solutions: Targeted guidance for specific symptoms, including migration failures, stuck VMs, quorum issues, storage errors, and configuration issues.
  • Data collection: Instructions for gathering diagnostic data if you have to contact Microsoft Support.

Before you begin, make sure that you have administrative access to the cluster resources, Hyper-V hosts, and management tools such as Failover Cluster Manager and Hyper-V Manager.

Troubleshooting checklist

Before you start to troubleshoot the issue, follow this checklist to help isolate the cause. The steps in this checklist might also help you fix the most common underlying issues.

Step 1: Gather information and review the symptoms

  • Gather and review log data. Review log entries for relevant events or messages that occurred near the time of your issue. As a starting point, review the following information:

    • System and Administration event log

    • Hyper-V event log (in Event Viewer, go to Applications and Services Logs > Microsoft > Windows > Hyper-V)

    • Any process dump files that you collected

    • Any network trace data that you collected

    • Cluster logs from all nodes
      To collect logs, run the following command on one of the cluster nodes:

      Get-ClusterLog -UseLocalTime -Destination <FolderPath>
      
  • Check the change and maintenance history. Check for any recent changes or operational events, including the following events:

    • Operating system updates
    • Driver or firmware changes
    • Configuration changes
    • Third-party software installations or updates
    • Scheduled maintenance
    • Cluster failover tasks
    • Storage migrations that affect VMs
  • Check host and VM versions: Make sure that the following components are up to date, and that the versions are compatible with one another:

    • Operating system of the Hyper-V host
    • VM configuration
    • Integration services

Step 2: Check the health of the cluster and the underlying infrastructure

  • To check the health of the failover clusters and cluster resources, use Failover Cluster Manager:

    • Identify any error messages in Failover Cluster Manager.
    • Identify which nodes are online and which, if any, are offline.
    • Identify any cluster resources that are stuck in Locked, Paused, or Not responding states.
    • Check the status of the CSV.
    • Make sure that the nodes can connect to the shared storage.
  • To check the health of the storage subsystem, run chkdsk in scan mode. Open a Windows Command Prompt window as an administrator, and then run the following command:

    chkdsk /scan
    

    If chkdsk identifies issues, see Data corruption and disk errors troubleshooting guidance.

  • Review the node network connections. Make sure that the following configurations are consistent on all nodes:

    • IP addresses
    • Network adapters
    • Teaming

    For more information, see Network recommendations for Hyper-V in a failover cluster.

    [!IMPORTANT]
    Network recommendations for Hyper-V in a failover cluster mentions load balancing and failover (LBFO) NIC Teaming. However, this technology is deprecated in favor of Switch-Embedded Teaming (SET). Whenever possible, use SET instead of NIC Teaming.

  • Make sure that any network device drivers support clustering, and that they're up to date.

Step 3: Check resource availability

For both the Hyper-V hosts and the guest VMs, use Task Manager and Performance Monitor (PerfMon) or other monitoring tools to assess the resource loads in the physical memory, CPU, storage, and network resource categories. Make sure that resources are consistently available. For more information about resource usage and Hyper-V, see Performance Tuning Hyper-V Servers.

Step 4: Check the cluster node configurations

Make sure that each cluster node has the same hardware and configuration, including the following components:

  • BIOS and firmware
  • Windows Server version and edition
  • CPU and memory

Step 5: Check the VM settings

  1. Make sure that the VMs are configured correctly to run on the cluster. For example, you can run the following command in a Windows PowerShell window on a cluster node:

    Get-ClusterResource | Where-Object { $_.ResourceType -eq "Virtual Machine" }
    

    For more information about how to configure VMs for replication in different types of topologies, see Replicate a virtual machine.

  2. Check the locations of the VM configuration files (.vmcx) and disk files (.vhd or .vhdx).

    • Make sure that the files reside on the shared storage that the cluster nodes use.
    • Make sure that your VM management client (such as Hyper-V Manager) uses the correct paths for these files.
  3. Check the VMs for checkpoints (also known as snapshots). Consider merging any existing checkpoints. If any checkpoints don't merge, see Hyper-V snapshots, checkpoints, and differencing disks (AVHDX) troubleshooting guidance.

  4. Make sure that the VM configurations are up to date, and that all host nodes support the VM configuration version.

Common issues and solutions

VMs don't migrate

The most common reasons that VMs can't migrate from one cluster node to another involve differences in the node configurations, and VM configurations that aren't supported in the same manner on all nodes. The following list links to detailed descriptions and resolution steps for specific symptoms:

VM or host becomes "stuck" and unresponsive

Symptoms
  • A VM becomes stuck in a transitional state (Starting or Stopping) or in the Offline state.

  • A VM or host enters one of the following states:

    • Paused
    • Critical
    • Not responding
    • Resource-locked
  • You might receive error messages that resemble the following examples:

    • Resource lock
    • Service unable to pause
    • Storage inaccessible
  • The event logs contain events that indicate service issues or timeouts, such as the following examples:

    • Event ID 153
    • Event ID 5120
Possible causes

This kind of issue can have several different causes. Typically, the messages and events that you observe indicate which cause is involved. Possibilities include the following causes:

  • A resource, such as a storage device, disconnected from the network.
  • A resource lock wasn't released correctly because an operation failed.
  • Hardware failed and reduced available resources (CPU, memory, storage IO), or resources are otherwise exhausted.
  • Drivers or firmware are out of date.
Resolution

To isolate the kind of issue that you're experiencing and address the most common issues, complete the Troubleshooting checklist. If the issue persists, use one of the following methods (depending on the kind of issue):

Cluster database or quorum issues

Symptoms
  • Cluster services don't start, and cluster doesn't come online.

  • Cluster goes offline intermittently.

  • CSV or VMs are offline or inaccessible.

  • You receive error messages or events that refer to either of the following issues:

    • Corrupted cluster database (CLUSDB)
    • Lost quorum
Possible causes
  • A failed operation corrupted the cluster database. Such operations include adding or removing a node, and shutting down the cluster.
  • The quorum disk or cluster witness is lost or inaccessible.
  • Firewall rules block communication between the cluster nodes.
Resolution

To isolate the kind of issue that you're experiencing and address the most common issues, complete the Troubleshooting checklist. For detailed checklists and troubleshooting information, see the following articles:

If you can't resolve the issue, you can, as a last resort, back up the cluster, validate the backup, and then rebuild the cluster.

Storage alerts, errors, or timeouts

Symptoms
  • The event logs contain events that indicate service issues or timeouts, such as the following examples:

    • Event ID 153
    • Event ID 157
    • Event ID 5120
  • You receive error messages or alerts about storage hardware (such as RAID or SAN devices)

  • Management tools label volumes as read-only, offline, missing, disconnected, or inaccessible

  • VMs unexpectedly restart, become unresponsive, or don't migrate

Possible causes
  • Storage drivers or firmware are faulty or outdated.
  • Disks were removed incorrectly.
  • A SAN failed over, or storage hardware malfunctioned
  • Storage paths or partitions aren't configured correctly.
Resolution

To isolate the kind of issue that you're experiencing and address the most common issues, complete the Troubleshooting checklist. If the issue persists, use one of the following methods (depending on the kind of issue):

Communication issues between cluster resources

Symptoms
  • Nodes drop from the cluster or enter quarantine, or heartbeats are missed or dropped.
  • RDP or discovery fails.
  • VMs migrate unexpectedly or don't migrate, particularly after nodes are moved or maintained.
Possible causes
  • Firewall rules are blocking required ports.
  • Duplicate IP addresses are configured.
  • DNS isn't configured correctly.
  • A network interface failed.
  • The cluster network isn't configured correctly (for example, client communication uses the wrong mode).
  • Teaming or VLAN settings aren't configured correctly.
Resolution

To isolate the kind of issue that you're experiencing and address the most common issues, complete the Troubleshooting checklist. If the issue persists, use one of the following methods (depending on the kind of issue):

  • Use Failover Cluster Manager to make sure that the different types of cluster network traffic (management, cluster, and client) follow the correct routing.

  • Make sure that all required ports are open. For more information about the required ports, see Hyper-V service and Cluster service in Service overview and network port requirements for Windows.

  • Verify that IP addresses are static.

  • Verify the DNS configuration, and resolve any conflicts. Make sure that no services make unexpected updates to DNS records.

VMs missing, not starting, or generating configuration error messages

Symptoms
  • After you restore or clone a VM, it won't start.
  • Your management client doesn't include a particular VM.
  • You can't import or export a VM.
  • You receive error messages that refer to Incomplete VM configuration.
  • The event logs contain Event ID 1069.
Possible causes
  • VM configuration files are stored on local (nonshared) storage.
  • Disk files or checkpoints are missing or orphaned, or the checkpoint merge process didn't complete correctly.
  • Configuration versions don't match.
Resolution

To isolate the kind of issue that you're experiencing and address the most common issues, complete the Troubleshooting checklist. If the issue persists, try the following methods:

Data Collection

If you need additional assistance, collect the following data, and then contact Microsoft Support.

General data

  • Version information (and build information, if applicable) for your infrastructure, including the following components:

    • Windows Server
    • Hyper-V
    • Integration services
    • Storage devices, firmware, and drivers
  • Screenshots of unexpected behavior or error messages

  • Trace or dump file data, including data files from ProcDump, ProcMon, PerfMon, or livekd

  • RAID and SAN logs

  • Documentation of recent changes and maintenance for each cluster node

  • Output of PowerShell queries that documents the state of the cluster, network, and storage systems

Data for Hyper-V related issues

Use Microsoft's troubleshooting scripts (TSS) to automatically gather information for troubleshooting. For instructions, see Collect data to analyze and troubleshoot Hyper-V scenarios.

Additionally, gather the following information:

  • Relevant events from the System and Application logs, and events from the Hyper-V log. The Hyper-V log is available in Event Viewer at Applications and Services > Microsoft > Windows > Hyper-V. You can export this information from Event Viewer.

  • VM configuration information. Use Hyper-V Manager or PowerShell to export this information.

Data for cluster-related issues

Use Microsoft troubleshooting scripts (TSS) to automatically gather information for troubleshooting. For instructions, see Collect data to analyze and troubleshoot clustering and high availability scenarios.

Additionally, gather the following information:

  • Cluster logs from all nodes To collect logs, run the following command on one of the cluster nodes:

    Get-ClusterLog -UseLocalTime -Destination <folder path>
    
  • Relevant event logs from the FailoverClustering and DNS Server event log channels

  • Details of the CNO object in Active Directory, including security permissions

References

Failover clusters

Hyper-V and virtual machines

PowerShell cmdlets

Storage

Windows Server

Source: Microsoft Learn / IT Pro troubleshooting by Microsoft and contributors. Documentation licensed CC BY 4.0. Code samples are covered by the MIT License. Converted from Markdown, with links normalized and embedded media omitted.

资料来源与适用范围

资料来源
Microsoft Learn / IT Pro troubleshooting
适用范围
- Supported versions of Windows Server · Source date 03/23/2026
内容语言
中文;命令、路径和错误标识保留技术原文
本地收录日期
2026-11-15
授权说明
CC BY 4.0 (documentation), MIT (code samples)

正文与所需资源由本站本地存储。查看来源及更新记录 ↗

请按适用版本和实际日志确认原因;本站没有对每一种系统或设备逐项执行文中的操作。