大网数据广告活动
故障转移群集

Troubleshoot Failover Cluster Live Migration Freeze with TSM Role

· 52CE 我爱测

Troubleshoot live migration hangs in Windows Failover Clusters, identify the TSM backup role as the root cause, and apply a safe fix during patching cycles.


Troubleshoot failover cluster live migration freeze with TSM role

Summary

A Windows Failover Cluster might hang and become unresponsive during live migration operations, such as role or VM movement between cluster nodes. This problem can occur intermittently and is often reproducible during scheduled patching cycles when you evacuate nodes. Use this article to resolve cluster hangs that occur during live migration when a third-party backup agent runs as a cluster role.

Symptoms

You might observe one or more of the following symptoms:

  • The failover cluster hangs or freezes during live migration of cluster roles.
  • Cluster nodes become unresponsive mid-migration.
  • The problem recurs consistently during patching windows.
  • No successful role movement occurs without the hang.

Memory dump analysis typically reveals the following findings:

  • A consistent deadlock pattern occurs when \Device\NetBt_If1 is being destroyed during the live migration process.
  • The Sysmon driver isn't updated for an extended period (for example, approximately five years). Outdated Sysmon drivers are a known source of deadlocks and network-related instability during live migration.
  • The PktMon.sys driver shows an old timestamp (for example, 2007-06-16), which indicates it's outdated.

Cause

The cause is the TSM (IBM Tivoli Storage Manager) backup service running as a Failover Cluster role. During live migration, the TSM cluster role interferes with the network teardown sequence, causing a deadlock.

Note: Although dump analysis might initially point to Sysmon and PktMon as potential contributors, the actual root cause is the TSM backup service role in the cluster. Disabling this role resolves the problem without any Sysmon or PktMon changes.

Resolution

Disable the TSM backup service role before performing live migration. You can apply the fix during a patching cycle.

  1. Identify all cluster roles and note any third-party backup agents, such as TSM or IBM Spectrum Protect, running as cluster roles.
  2. In Failover Cluster Manager, disable the TSM backup service role before starting live migration or node evacuation.
  3. Perform the live migration or node drain operation.
  4. Re-enable the TSM backup service role after the migration finishes and the node is back online.

Important: Don't remove or delete any built-in Windows drivers, such as PktMon.sys, from production environments without thorough testing in a nonproduction environment first. PktMon is a built-in Windows network traffic analyzer and should only be updated, not removed.

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 05/12/2026
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。