无法启动分发。错误代码： 16，失败步骤： 2 错误代码: Wsl/Service/CreateInstance/E_FAIL

### Windows Version

Microsoft Windows [版本 10.0.26100.7462]

### WSL Version

WSL 版本: 2.6.3.0

### Are you using WSL 1 or WSL 2?

- [x] WSL 2
- [ ] WSL 1

### Kernel Version

Linux version 6.6.87.2-microsoft-standard-WSL2 (root@439a258ad544) (gcc (GCC) 11.2.0, GNU ld (GNU Binutils) 2.37) #1 SMP PREEMPT_DYNAMIC Thu Jun 5 18:30:46 UTC 2025

### Distro Version

Ubuntu 24.04

### Other Software

_No response_

### Repro Steps

我在尝试实现[Open-AutoGLM](https://github.com/zai-org/Open-AutoGLM/)中的“如果用户选择选项 B(本地部署模型)”，使用wsl部署，执行的命令如下，能成功部署，再对服务进行ctrl+C 中断和不关闭服务的两种情况下，我使用了“wsl --shutdown” 或者是在linux中“ shutdown”之后再执行`wsl -d Ubuntu`均会出现
```
(base) PS E:\> wsl -d Ubuntu
无法启动分发。错误代码： 16，失败步骤： 2
错误代码: Wsl/Service/CreateInstance/E_FAIL
```

具体的安装linux 中项目安装

```pwsh
wsl --unregister Ubuntu
cd E:
wsl --import Ubuntu E:\WSL\Ubuntu2404 E:\ubuntu-24.04.3-wsl-amd64.gz --version 2

wsl -d Ubuntu

```

``` bash
mkdir -p ~/miniconda3
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda3/miniconda.sh
bash ~/miniconda3/miniconda.sh -b -u -p ~/miniconda3
rm ~/miniconda3/miniconda.sh
```
``` bash
source ~/miniconda3/bin/activate
conda init bash
```
```bash
conda create -n autoGLM python=3.11 -y
conda activate autoGLM
cd /mnt/d/GLM/Open-AutoGLM
pip install torch==2.9.0 torchvision==0.24.0 torchaudio==2.9.0 --index-url https://download.pytorch.org/whl/cu128
pip3 install -r requirements.txt
pip3 install -e .
pip install -U transformers

```
```bash
apt update
apt install -y build-essential
apt install net-tools
netstat -tulpn
```
```bash
python3 -m vllm.entrypoints.openai.api_server \
--served-model-name autoglm-phone-9b \
--allowed-local-media-path / \
--mm-encoder-tp-mode data \
--mm_processor_cache_type shm \
--mm_processor_kwargs "{\"max_pixels\":5000000}" \
--max-model-len 25480 \
--chat-template-content-format string \
--limit-mm-per-prompt "{\"image\":10}" \
--model zai-org/AutoGLM-Phone-9B \
--host 0.0.0.0 \
--port 5050
```

### Expected Behavior

我能正确的启动wsl,而不是每次都要重新安装机器学习环境。

### Actual Behavior

[WslLogs-2026-01-10_02-34-42.zip](https://github.com/user-attachments/files/24533701/WslLogs-2026-01-10_02-34-42.zip)

### Diagnostic Logs

_No response_

Start by reviewing WslLogs-2026-01-10_02-34-42.zip and the reproduction commands using `wsl --shutdown` followed by `wsl -d Ubuntu`. Compare the behavior after importing the Ubuntu 24.04 distribution and running the Open-AutoGLM service. Done means the existing WSL installation starts normally without reinstalling its machine-learning environment.

Written by the indexing model from the issue text.