I've run some experiments with hive on tez. I have a query that 80% of the time runs for about 45 seconds and the rest of the time runs for about 24 seconds. What could that rare speedup be attributed to?
Asked
Active
Viewed 29 times
0
pavel_orekhov
- 1,657
- 2
- 15
- 37
-
are you measuring the time by executing query on dedicated queue, if not please do so. – Gaurang Shah Feb 05 '19 at 15:35
-
@GaurangShah how? In Yarn? And I run queries in a sandbox, so it probably should not matter? – pavel_orekhov Feb 05 '19 at 15:38
-
it shouldn't. however I don't think sandbox is the best to measure the time. if you are using docker your resources will be shared with other containers and host machine. if you are using VM, I think it would be more appropriate – Gaurang Shah Feb 05 '19 at 15:59
-
@GaurangShah yes, I run it inside virtual box. – pavel_orekhov Feb 05 '19 at 16:01