Hello Justin, Thanks for your answer. I confirm my cluster is formed properly (AFAIK) and the failover is handled on the cluster side.
I uploaded the following elements on this gist: https://gist.github.com/alexandre-touret/3066f451161ce48243d24286 - Primary broker.xml configuration - Replica broker.xml configuration - Primary logs during a failover - Backup logs during a failover - Client pom.xml, application.properties & Java code Thanks in advance for your help. Best Regards, Alexandre ---------- Forwarded message --------- From: Justin Bertram < jbert...@apache.org > Date: Thu, 4 Sept 2025 at 20:57 Subject: Re: Unable to handle failover in a Spring client with the CORE protocol To: < users@activemq.apache.org > Can you confirm that your cluster is forming properly and when the primary is stopped the backup activates? Do you have details on the specific exception that the client is seeing? Can you provide the actual, full URL that you're using (i.e. with the properties expanded/replaced)? Lastly, can you confirm that the XML you pasted is, in fact, from your primary node? It looks to me like it's from the backup. Justin On Thu, Sep 4, 2025 at 10:40 AM < alexandre.tou...@free.fr .invalid> wrote: > Hello, > I'm currently trying to set up a Spring Artemis client and particularly > the failover mechanism. > > I sat up Artemis using a cluster two nodes replicated using a static > definition: > > Below my primary node configuration (from the broker.xml file) > > <cluster-user>USER</cluster-user> > <cluster-password>PASSWORD</cluster-password> > <ha-policy> > <replication> > <backup> > <allow-failback>true</allow-failback> > </backup> > </replication> > </ha-policy> > <connectors> > <connector name="backup">tcp://artemis-backup:61616</connector> > <connector name="primary">tcp://artemis-primary:61616</connector> > </connectors> > <acceptors> > <acceptor name="artemis">tcp:// 0.0.0.0:61616?protocols=OPENWIRE,CORE > </acceptor> > </acceptors> > > <cluster-connections> > <cluster-connection name="my-cluster"> > <connector-ref>backup</connector-ref> > <static-connectors> > <connector-ref>primary</connector-ref> > </static-connectors> > <retry-interval>1000</retry-interval> > <use-duplicate-detection>true</use-duplicate-detection> > </cluster-connection> > </cluster-connections> > > > On the client side, I use Spring Boot & the artemis starter > > <dependency> > <groupId>org.springframework.boot</groupId> > <artifactId>spring-boot-starter-artemis</artifactId> > </dependency> > > and configured the properties as following: > > > spring.artemis.broker-url=(${ARTEMIS_PRIMARY_URL}?name=primary,${ARTEMIS_BACKUP_URL}?name=backup)?protocol=CORE&ha=true&reconnectAttempts=1000&failoverAttempts=1000&clientFailureCheckPeriod=30000 > > spring.artemis.user=USER > spring.artemis.password=PASSSWORD > spring.artemis.mode=native > > Unfortunately, when I run several write/read calls using the JMSTemplate, > it fails when the primary node stops and the backup node takes over it. An > exception is automatically raised and the client doesn't automatically > shift to the backup server on the fly. > Otherwise, when I re-try my request, it's OK. > > When I try to do the same with the OPENWIRE protocol and the failover URL, > it works. During the same thread, the failover is perfectly handled. > > How can I get the same behaviour with the CORE protocol? > > Thanks in advance. > > Best Regards, > Alexandre > > --------------------------------------------------------------------- > To unsubscribe, e-mail: users-unsubscr...@activemq.apache.org > For additional commands, e-mail: users-h...@activemq.apache.org > For further information, visit: https://activemq.apache.org/contact > > > Alexandre Touret --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@activemq.apache.org For additional commands, e-mail: users-h...@activemq.apache.org For further information, visit: https://activemq.apache.org/contact
