A Simple Way to Run One Copy of a Job on Many Servers

In many school projects or small applications, it is common to run operations such as sending a report or cleaning old data regularly. If you have only one server, you can handle this efficiently with a cron job. However, if you have multiple servers, operating the same job simultaneously would be time-consuming. This paper illustrates a simple solution: let all servers “race” to insert one row into a normal database table. The database’s unique constraint makes sure only one succeeds. The successful server performs the operation. After it finishes, it removes the row so the process can be performed repeatedly. We do not need special tools or external help; just one small table and a simple trick.