Bash exit status在脚本中失效：Apache启停脚本错误提示异常求助

Hey there! Let's tackle this problem where your stop action returns a success even when Apache isn't running. This is super common with service scripts, and it almost always boils down to how you're checking if Apache is actually active before running the stop command.

The Root Cause #

Most service commands like systemctl stop apache2 or service apache2 stop will return a "success" exit code (0) even if the service was already stopped. Your script is probably just running the stop command and assuming it worked, without verifying if Apache was actually running in the first place.

Step-by-Step Fix #

1. Add a Reliable Status Check First
   Before trying to stop Apache, you need to confirm it's actually running. The most reliable way (for systemd-based systems, which is most modern distros) is using systemctl is-active --quiet apache2:

   • Returns exit code 0 if Apache is running
   • Returns non-zero if it's stopped/failed

2. Update Your Stop Logic
   Modify your script's stop case to first check status, then act accordingly. Here's a sample implementation:

#!/bin/bash

case "$1" in
    start)
        # Check if Apache is already running
        if systemctl is-active --quiet apache2; then
            echo "Error: Apache is already running!" >&2
            exit 1
        fi
        # Try to start Apache
        if systemctl start apache2; then
            echo "Apache started successfully."
        else
            echo "Failed to start Apache." >&2
            exit 1
        fi
        ;;
    stop)
        # Check if Apache is NOT running first
        if ! systemctl is-active --quiet apache2; then
            echo "Error: Apache is not running - cannot stop it!" >&2
            exit 1
        fi
        # Try to stop Apache
        if systemctl stop apache2; then
            echo "Apache stopped successfully."
        else
            echo "Failed to stop Apache." >&2
            exit 1
        fi
        ;;
    status)
        if systemctl is-active --quiet apache2; then
            echo "Apache is running."
        else
            echo "Apache is stopped."
        fi
        ;;
    *)
        echo "Usage: $0 {start|stop|status}" >&2
        exit 1
        ;;
esac

3. Key Improvements in This Script
   • Pre-checks: We verify Apache's state before running start/stop commands, so we can throw accurate errors if the action doesn't make sense.
   • Exit Code Checking: We capture the exit code of the systemctl start/stop commands to handle cases where the service fails to start/stop even if it was in the right state.
   • Error Stream: We send error messages to stderr (>&2) instead of stdout—this is a best practice for scripts, as it separates normal output from errors.

For Non-Systemd Systems #

If you're on an older distro using SysVinit instead of systemd, replace the systemctl checks with this:

# Check if Apache is running
if service apache2 status | grep -q "active (running)"; then
    # Running logic
else
    # Stopped logic
fi

This should fix your false success issue—now when you run the stop command and Apache isn't running, you'll get a clear error message instead of a misleading success.

内容的提问来源于stack exchange，提问作者Hebruiser