Originally posted by: mrp
there has been discussions in the past on this topic but couldn't get a definitive answer.
is LUSE on HDS arrays a concat or a stripe ? in the manuals it says concatenation, but i remember from my earlier job at Sun, that HDS claims it is stripe although mentioned as concat.
this would be important for me since if i want to get more performance atleast for OLTP type applications, i would rather take individual LDEVs from each parity group and then stripe at OS level using volume manager which would enabe data to be written across many parity groups at the same time.
but if LUSE are in fact concat, then i might not get required performance as a stripe since concat if i understand correctly will mostly benefit sequential type applications.