[Bug Report] Seemingly DNS failure after a short interval of time

### Checks

- [X] I have searched the existing issues
- [X] I have read the documentation
- [X] Is it your first time sumbitting an issue

### Current Behavior

After a certain period of time, dae fails to dial with error:
```
dial to wss://SERVER: dial tcp: lookup SERVER: device or resource busy
```
However, the `dig SERVER` works just fine. Not sure if it's actually a DNS failure.

### Expected Behavior

_No response_

### Steps to Reproduce

My config:
```
global{
wan_interface: auto

log_level: info
allow_insecure: false
# auto_config_kernel_parameter: true
dial_mode: domain
}

# Forward all DNS requests to local DNS
dns {
upstream {
dcompass: 'udp://127.0.0.1:53'
}
routing {
request {
fallback: dcompass
}
}
}

node {
v2ray: "NODE DETAIL"
}

group {
first {
policy: fixed(0)
}
}

routing{
# Don't route dcompass's requests
pname(NetworkManager, dcompass) -> must_direct

domain(geosite:category-ads) -> block
dip(geoip:private) -> direct
dip(geoip:cn) -> direct
# domain(geosite:cn) -> direct

# Don't route games that have server in China.
domain(geosite:category-games@cn) -> direct

fallback: first
}
```

### Environment

- **Dae version (use `dae --version`)**: v0.7.0
- **OS (e.g `cat /etc/os-release`)**:
```
ANSI_COLOR="1;34"
BUG_REPORT_URL="https://github.com/NixOS/nixpkgs/issues"
BUILD_ID="24.11.20240804.cb9a96f"
DOCUMENTATION_URL="https://nixos.org/learn.html"
HOME_URL="https://nixos.org/"
ID=nixos
IMAGE_ID=""
IMAGE_VERSION=""
LOGO="nix-snowflake"
NAME=NixOS
PRETTY_NAME="NixOS 24.11 (Vicuna)"
SUPPORT_URL="https://nixos.org/community.html"
VERSION="24.11 (Vicuna)"
VERSION_CODENAME=vicuna
VERSION_ID="24.11"
```
- **Kernel (e.g. `uname -a`)**: `Linux x1c7 6.10.3 #1-NixOS SMP PREEMPT_DYNAMIC Sat Aug 3 07:01:09 UTC 2024 x86_64 GNU/Linux`
- **Others**:

### Anything else?

_No response_

Start by reproducing the failure with the provided dae configuration on the reported NixOS and kernel versions, then compare dae's WSS lookup behavior with `dig SERVER` after the interval. Trace the relevant DNS and dialing paths in the Go code; done means identifying the cause and preventing the recurring `device or resource busy` failure, with a regression test if the affected path supports one.

Written by the indexing model from the issue text.