How to avoid the duplicates at Queue level in a clustered environment?

Issue

• User is facing a problem in production environment while running Quartz timers,
• When ever an endpoint is not available/reachable user is placing the request in a queue for retry, for this user is using Quartz timers to pick failed requests but user has 3 instances clustered and the records are duplicated.
• This is causing to increase the Queue depth.