Sometimes some servers don't respond to MAM query, which causes holes in history/archive. I'd suggest adding retry mechanism (let's say 5-10 attempts with increasing delay) in case of timeout to maintain complete history.
Currently retry is done on reconnection and is "good enough". Maybe we would like to "retry" in some other cases or after a timeout without reconnection? Thing to consider for 5.0
(this is a followup to #issue #244)
Sometimes some servers don't respond to MAM query, which causes holes in history/archive. I'd suggest adding retry mechanism (let's say 5-10 attempts with increasing delay) in case of timeout to maintain complete history.