Can you describe more about your test itself? Were your IoT processes publishing to queues or topics? To a single one or to one per client? Was anything consuming the messages? Were the messages persistent or non-persistent? (If persistent, can you prove it by showing a reasonable rate of growth on the KahaDB data files?)
Tim On Jul 2, 2017 9:40 PM, "Hidekazu" <hi_de...@yahoo.co.jp> wrote: > Dear christopher.l.shannon > > Thanks for your quick reply. > > I alocated 3GB memory for ActiveMQ. > This is our startup setting for heap allocation. > > /usr/sbin/java -Xms3G -Xmx3G -XX:PermSize=256M -XX:MaxPermSize=256M > > And this is the systemUsage setting in activemq.xml > > > <systemUsage> > <systemUsage> > <memoryUsage> > <memoryUsage memoryUsage percentOfJvmHeap="70" /> > </memoryUsage> > <storeUsage> > <storeUsage limit="100 gb" /> > </storeUsage> > <tempUsage> > <tempUsage limit="50 gb" /> > </tempUsage> > </systemUsage> > </systemUsage> > > > Thanks! > Hidekazu > > > > -- > View this message in context: http://activemq.2283324.n4. > nabble.com/AcativeMQ-becomes-slow-suddenly-in-60-minutes- > after-starting-a-performance-test-tp4728103p4728185.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. >