Code review comment for lp:~pbeaman/akiban-persistit/fix-several-small-bugs

VolumeStorageV2#updateMetaData() - including timestamp changed in result will make meta data always "dirty" and then written. Is that what we want?

ExchangeTest#traverseEQfalse0 - The bug has has 'false' in the title and 'true' in the description. Could we add cases for 'true' as well? Additionally, the server use case probably did it inside of a transaction. It should work as is if another test method just calls the new one, but inside of a transaction block. I have no reason to suspect a failure, but simple to add.

Persistit#close() - _journalManager.stopCopier() is no longer called. Should it be? The rest of the related changes look plausible and the new test is good.