i am sorry, maybe my question is not important but i am confusing. maybe someone can answer my question.
i have tried to use hbase in hadoop. when i query and find a value inside Hbase with filter and i try until more than 5 times, why the time is always drop ? drop mean faster than before.
first time i try to scan, i get maybe 16 second
and then second time i get 20
but the third untl the sixth i get 16,14,13,... and the low limit of my query time is 12 second to get 853 rows .
anyone can explain me ? why hadoop or hbase like a machine learning who can faster if we try more than 1 time ?
thanks a lot.
