Same recurring job executes on all servers

I was under the impression that recurring jobs or any job is processed only once from one of the servers, rather than all processing at the same time.

Please see from logs that the same jobs are executed across servers at the same time almost:

Why is that? And how to prevent this?

And here is the definitions:

We cannot tolerate same jobs running across servers, especially at the same time.

Start by comparing the server logs with the recurring-job definitions shown in the issue, focusing on whether the same schedule is registered across servers. Review the project's recurring-job and multi-server processing behavior to identify the expected coordination mechanism. Done means explaining why duplicate execution occurs and documenting a verified way to prevent it.

Written by the indexing model from the issue text.