Hi Rahman- ActiveMQ 5.x includes duplicate message detection in both client and broker-side areas. For failover producers, there is the ability to track messages and take steps to flush or reduce duplicate during failover.
The broker contains audit points at several places (Destinations and KahaDB store) to look for message dupes (based on message ID) that accounts for several of the low-level and broker-to-broker scenarios. Additionally, when messages are paged out, dupe checking is available to make sure duplicate messages are not paged-in from the store (recent fix added a counter metric for this, instead of copying duplicate detected messages to the DLQ: https://issues.apache.org/jira/browse/AMQ-8397) [Client Connection Factory] setAuditDepth ref: https://activemq.apache.org/maven/apidocs/org/apache/activemq/ActiveMQConnectionFactory.html [Failover Transport] trackMessages ref: https://activemq.apache.org/failover-transport-reference.html [Destination Policies] enableAudit = true | false maxAuditDepth maxProducersToAudit ref: https://activemq.apache.org/per-destination-policies [KahaDB] maxFailoverProducersToTrack failoverProducersAuditDepth Are you able to share more specifics about your use case or messaging pattern? ‘Duplicate messages’ can be a tricky thing in distributed messaging (regardless of product used) — sometimes it is a resend of the same message (ie failover), sometimes it is a ‘redelivery’ to the consumer (ie. consumer rollback or does not ack properly). In clusters, you can have redundant publishing of messages (ie. Pub-sub storm) to account for. Thanks, Matt Pavlovich > On Jun 1, 2022, at 1:15 PM, Gunawan, Rahman (GSFC-703.H)[Halvik Corp] > <rahman.guna...@nasa.gov.INVALID> wrote: > > Hi, > Does ActiveMQ 5.15.4 have duplicate message detection? If yes, where can I > find the documentation? > > Thanks > > Regards, > Rahman
