How to remove this single point of failure from application architecture?

Question

I have an application that currently has the following setting:

• Replicated MySQL DB
• Distributed work queue
• Multiple Consumers / Consumer Work Queues
• One manufacturer that adds jobs to the queue (server bottom to bottom)

The setup looks something like this:

Manufacturer queries the database for new items that need to be added to the list of recurring jobs that need to be added to the work queue every N minutes. This job producer is the only node in my entire architecture that, if unsuccessful, will lead to the failure of the whole process. I can have a database server, a queue server, or several production servers, and the process will continue to work.

, ? , , , N . .

, 1/P , P - .

- :

itemsToBeProcess = db.FetchItems()
for (item in itemsToBeProcessed) {
    if item.id % producerNumber == 0) // Queue job
}

. , , , , 1/P, . , .

- - , ?

Answer 1

- db N ? ​​ , N N (, "open" → "in progress" ), " " * ( , ( ) ). / FO ..