Hi,
I was benchmarking some of the phoenix queries with different compaction level
tuning.
A strange thing is observed when there are huge number of Hfiles on disk. The
queries not returning any data (resultset size 0) execute very quickly (5-10 ms
or so) but just doing a rs.next() on result set returned from such a query
takes over 50 seconds! This is consistently reproducible.
I'd assume that when the prepared statement executes and prepared the result
set, it has completed reading the table or memstore and/ or Hfiles on disk, so
the time it takes to execute the query should have included that time. Hence,
when I do rs.next() on such a result set produced, it should not take huge time
esp. when it is empty.
What am I missing or has anyone else faced/ solved this issue?
Thanks,Sumit
