well....after some profiling (and reading the docs), I found an explanation on my own:-)
The mock endpoints used in my unit test are the reason for the memory increase. The mocks keep all messages in memory, just as it is documented. I just did not see it in the docs. So if I change my route to use e.g. a dataset endpoint instead of the mocks, the aggregator works fine with a constant memory footprint. Maybe this info is helpful for somebody else... cheers Klaus -- View this message in context: http://camel.465427.n5.nabble.com/Aggregation-performance-depending-on-heap-space-tp5114812p5114862.html Sent from the Camel - Users mailing list archive at Nabble.com.
