如何从bash脚本执行的bash脚本返回PID

下面的get_pid（）函数用于返回

守护进程的pid。sh

下面的脚本与

daemon\u-tinerary.sh

不在同一工作目录中

#!/bin/bash

PID=""

get_pid() {
    PID='pidof daemon_itinerary.sh'
}

start() {
    echo "Restarting test_daemon"
    get_pid
    if [[ -z $PID ]]; then
        echo "starting test_daemon .."
        sh /var/www/bin/daemon_itinerary.sh &
        get_pid
        echo "done. PID=$PID"
    else
        echo "test_deamon is alrady running, PID=$PID"
    fi
}

case "$1" in
start)
    start
;;
...
*)
    echo "Usage: $0 {start|stop|restart|status}"
esac

*编辑

---

Start作为命令行参数传入

我们使用pgrep获得如下流程的pid

PID=$(pgrep -f "daemon_itinerary.sh" | xargs)

# xargs - is given because pgrep will return both process id as well as parent pid
# also it will help us to get all pids if multiple instances are running.
# pgrep option to get session id or parent id alone, here its from manual
# -P, --parent ppid,...
#     Only match processes whose parent process ID is listed.
# -s, --session sid,...
#     Only match processes whose process session ID is listed. Session ID 0 is translated into pgrep's or pkill's own session ID.

---

那么，问题是什么-|您不是从任何地方调用start（）吗？pidof daemon_Cinerary.sh应该是反勾，而不是单引号。使daemon_Cinerary.sh可执行（700）并直接调用，而不是使用“sh”。然后使用'pidof-x daemon_-itinery.sh'实际上不是，正如Roadawl建议的那样，您应该使用pidof-x daemon_-itinery.sh。要在变量中捕获PID，它应该是PID=$（pidof daemon_tinerary.sh）。请注意pid变量的小写字母，因为pid在bash中是保留名称。