macOS 10.13下LaunchDaemon脚本无法发送邮件的问题求助

Got it, let's break down why your script works in the foreground but fails when run via LaunchDaemon—this almost always comes down to environment differences between interactive shells and the stripped-down launchd runtime. Here's how to fix it step by step:

1. First, Capture Debug Logs #

LaunchDaemon doesn't show output by default, so let's add logging to your plist to see exactly what's breaking. Update your .plist file in /Library/LaunchDaemons to include log paths:

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
    <key>Label</key>
    <string>com.yourname.vmcheck</string> <!-- Use a unique label -->
    <key>ProgramArguments</key>
    <array>
        <string>/bin/bash</string>
        <string>/full/path/to/your/script.sh</string> <!-- Always use absolute paths! -->
    </array>
    <key>StartInterval</key>
    <integer>60</integer>
    <!-- Add logging -->
    <key>StandardOutPath</key>
    <string>/var/log/vmcheck_out.log</string>
    <key>StandardErrorPath</key>
    <string>/var/log/vmcheck_err.log</string>
    <key>RunAtLoad</key>
    <true/>
</dict>
</plist>

Then reload the daemon and check logs after a minute:

sudo launchctl unload /Library/LaunchDaemons/com.yourname.vmcheck.plist
sudo launchctl load /Library/LaunchDaemons/com.yourname.vmcheck.plist
cat /var/log/vmcheck_err.log

This will tell you if commands like vmrun or sendmail are missing, or if Postfix is throwing errors.

2. Fix Path & Environment Variables #

LaunchDaemon runs with a minimal PATH (usually just /bin:/usr/bin), which means commands installed in non-standard locations (like vmrun inside VMware Fusion's bundle) won't be found. Fix this two ways:

Option A: Use Absolute Paths in Your Script #

Update every command in your script to use full paths:

#!/bin/bash
# Full path to vmrun
VM_RUN="/Applications/VMware Fusion.app/Contents/Library/vmrun"
# Full path to your VM's .vmx file
VM_PATH="/Users/yourusername/Documents/Virtual Machines/Ubuntu/Ubuntu.vmx"

# Check if VM is running
if ! "$VM_RUN" list | grep -q "$VM_PATH"; then
    # Use absolute path to sendmail
    /usr/sbin/sendmail -t <<EOF
To: your-alert-email@example.com
From: vm-monitor@your-mac.local
Subject: Ubuntu VM is Down!

Your VMware Fusion Ubuntu VM has stopped running. Please log in to restart it.
EOF
fi

Option B: Explicitly Set PATH in the Plist #

Add an EnvironmentVariables section to your .plist to expand the available paths:

<key>EnvironmentVariables</key>
<dict>
    <key>PATH</key>
    <string>/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/Applications/VMware Fusion.app/Contents/Library</string>
</dict>

3. Verify Postfix & Permissions #

Since your script works in the foreground, Postfix is configured correctly for your user—but LaunchDaemon runs as root by default. Make sure:

• Postfix's system-wide configuration (in /etc/postfix/main.cf) includes your relay settings, not just user-specific configs.
• The root user has permission to send mail through your relay. If your relay requires SASL auth, ensure the credentials are stored in a system-wide Postfix config (not just ~/.postfixrc for your user).

4. Validate Plist Permissions #

LaunchDaemon requires strict file permissions for plists:

sudo chown root:wheel /Library/LaunchDaemons/com.yourname.vmcheck.plist
sudo chmod 644 /Library/LaunchDaemons/com.yourname.vmcheck.plist

Incorrect permissions will cause launchd to ignore the plist entirely.

Final Test #

After making these changes, reload the daemon and wait a minute. Check the logs again—if everything works, you should see no errors in vmcheck_err.log, and get your alert email when the VM is down.

内容的提问来源于stack exchange，提问作者stubert