impala in SQL optimization methods

1. fetch water table, if data is to use all partitions can not access data from the layers SA, from the need to change the number of layers taken SH, SH layer parquet as storage, better query performance.


2. For the script used in temporary tables, if the following conditions are required information tables
    1) a larger amount of data per se
    2) requires a large amount of data and the associated table
    3) itself be used multiple times more


3. For the calculation of repeated use SQL, need to calculate in advance the data into a temporary table, saving computing resource consumption.


4. The left section associated with the least possible SQL join the like can be performed multi-block some SQL.

Guess you like

Origin www.cnblogs.com/hello-wei/p/11883810.html