The test is running slow after changing defaults on block and sync options.
The test is sending and consuming 1000 messages on every iteration. Making it to block on either sending or consuming would mean to wait the callback and roundtrip from every message sent, and every message consumed.
I didn't find any other problems besides the sync changes.
I am now changing those parameters for that test, as this test really need block=false, and sync=false, as it is validating the concurrent compact of the journal.