Mastering Shell Script Timeouts with the Linux timeout Command

In shell scripting, scripts can sometimes hang, leading developers to use crude watchdog loops that may kill the wrong process when new background jobs start.

To avoid indiscriminate termination, the built‑in #man timeout command (available on CentOS 6) should be used.

For example, to ping www.baidu.com but stop after 5 seconds of no response, run: #timeout 5s ping www.baidu.com The command exits after five seconds, as shown in the following output:

Building on this, the author improves a previously written MQ watchdog script by adding timeout monitoring and email alerts for failures. The original script took about 13 seconds to run; a timeout of 20 seconds is set, and the execution result is displayed:

When the script finishes normally it returns code 0; if it is terminated by the timeout, it returns code 124 (while a manual Ctrl+C yields 130). This behavior is illustrated below:

The final script (shown as an image) can be placed in crontab for scheduled execution.

Additional notes:

Using #timeout 3s ping www.baidu.com is not ideal because the Unix ping command repeats indefinitely; a better test is #timeout 3s sleep 1 versus #timeout 3s sleep 5 to observe timeout behavior.

The timeout command does not work when its argument is a shell function.