I'm pretty sure there's no other way than to tell producers to switch to persistent mode. http://activemq.apache.org/how-do-i-make-messages-durable.html
Only way to force persistence would be to have a custom plugin that alters incoming messages. That would be considered jms contract breaking, but if your producers can't be altered might be the way to go. Also make sure that persistence is enabled in activemq.xml. Highly unlikely you have this overwritten and it is on by default, but such stuff happens sometimes. http://activemq.apache.org/how-do-i-disable-persistence.html pt., 12 paź 2018, 08:43 użytkownik Mohsin Khan <mohsin.khan5...@gmail.com> napisał: > I have setup a Redhat HA cluster for ActiveMQ. > We have 2 nodes and one virtual ip over which clients are accessing the > broker. > > I am using KahaDB, available over nfs to both the servers. The setup is > master-slave, once master goes down the slave is taking from there but the > main issue is the queues and the topic the messages are getting clear from > them whever the flipping is happing from master to slave. > > Do i need to add or modify anything in activemq.xml so that the messages > remain persistent over restarts or i have to make aware of the producers to > send persistent messages. > > If you have any other suggestion for ActiveMq clusters in production, do > share. > > Regards > Mohsin > System Administrator-Linux > > > > -- > Sent from: > http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html >
