在虚拟机中运行Windows Docker触发崩溃的问题排查求助

Got it, let's walk through actionable steps to diagnose this frustrating issue where your Windows Server 2016 VM crashes with an unhandled exception whenever you run a Docker container. Here's what you can check:

1. Validate Resource Allocation (Host & VM) #

Windows containers are resource-heavy, so insufficient resources often trigger crashes:

• Ensure your VM has at least 4GB of RAM (8GB+ recommended) and 2+ CPU cores allocated. Windows Server 2016 + Docker needs this baseline to run reliably.
• Monitor the host (Windows Server 2012) resource usage in Task Manager while you attempt to start the container. If the host hits 100% CPU or memory, that could starve the VM and cause a crash.

2. Verify Docker-Windows Version Compatibility #

Mismatched versions are a common culprit for stability issues:

• Windows Server 2016 only supports specific Docker releases. Stick to Docker EE (Enterprise Edition) stable builds compatible with 2016—like 18.09.x or 20.10.x (avoid bleeding-edge versions).
• Test with a minimal, trusted container first to rule out your target image:

  docker run mcr.microsoft.com/windows/nanoserver:1607
  

  If this works, the problem is likely tied to your specific container image, not Docker itself.

3. Capture & Analyze VM Crash Data #

Since Docker logs are corrupted, we need to look at system-level crash info:

• Enable crash dumps on the VM: Right-click "This PC" → Properties → Advanced System Settings → Startup and Recovery → Settings. Choose "Small Memory Dump" or "Core Memory Dump" and set a save path. After the next crash, use tools like WinDbg to analyze the dump file—it’ll show which process triggered the unhandled exception.
• Check the VM’s System Event Log: Open Event Viewer → Windows Logs → System, filter for "Error" and "Critical" events around the crash time. Look for disk errors, driver failures (especially virtualization-related drivers), or Hyper-V integration service issues.

4. Fix Corrupted Docker Logs #

Before troubleshooting further, get Docker logging working again:

• Stop the Docker service first:

  Stop-Service Docker
  

• Backup the corrupted logs (usually stored at C:\ProgramData\Docker\logs) then delete the damaged files.
• Restart Docker:

  Start-Service Docker
  

• If logs get corrupted again immediately, check the VM’s disk health. Run chkdsk C: /f in the VM (you’ll need to restart to schedule the scan) and verify the host’s virtual disk isn’t corrupted via Hyper-V Manager.

5. Check Host Virtualization Configuration #

Hyper-V misconfigurations can cause VM instability:

• Confirm the host (Windows Server 2012) has Hyper-V enabled and that hardware virtualization (Intel VT-x/AMD-V) is turned on in the BIOS/UEFI.
• Ensure Hyper-V Integration Services are up-to-date on both the host and VM. Mismatched versions can lead to communication issues between host and guest.

6. Isolate the Container Image as the Root Cause #

If minimal containers work fine, focus on your target image:

• Try pulling a different version of the target image (e.g., if you’re using my-image:latest, try my-image:v1.0).
• Run the same image on a separate, working Windows Server 2016 environment. If it crashes there too, the image itself has a bug (like a faulty application inside the container triggering an unhandled exception that propagates to the VM).

内容的提问来源于stack exchange，提问作者Jay Jay Jay