I was using GlusterFS to keep my master/slave shared file system in sync. When embedded broker clients disconnected from the stand alone broker somehow the indexes were getting corrupted and maybe some data files too.
I tried the same test several times without Gluster keeping a single master broker, no slave (because GlusterFS was my distributed FS mechanism) and never saw these issues again. I blame GluterFS for these issues, probably there's some tuning to be done but personally I had a bad experience with it and would avoid it for medium load. -- View this message in context: http://activemq.2283324.n4.nabble.com/Broker-stuck-showing-error-messages-in-log-tp4656612p4657222.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
