hive mapjoin 方案

hive> set  hive.auto.convert.join=true;

hive> set  hive.auto.convert.join;
hive.auto.convert.join=true

hive> SELECT 
    -- 可以显示的指定如下这一行mapjoin 关键词。
    /*+mapjoin(b)*/
    a.date,a.page_id,b.page_name,
    count(1) as pv,
    count(distinct gu_id) as uv
    from dw.fct_page_ref a
    left join dw.dim_page b on a.page_id = b.page_id 
    where date = "2017-09-20"
    group by a.date,a.page_id,b.page_name;

Query ID = gongzi_20170922102424_c2e3e956-1c3f-41de-86b6-8531b57a23ab
Total jobs = 1
Execution log at: /tmp/gongzi/gongzi_20170922102424_c2e3e956-1c3f-41de-86b6-8531b57a23ab.log
2017-09-22 10:25:01     Starting to launch local task to process map join;      maximum memory = 1908932608
....
...