| --! qt:dataset:src |
| set hive.merge.mapredfiles=true; |
| set hive.auto.convert.join=false; |
| set mapreduce.job.reduces=1000; |
| |
| create table dummy_n3 (a string); |
| insert overwrite directory '/tmp/test' select src.key from src join dummy_n3 on src.key = dummy_n3.a; |
| dfs -ls /tmp/test; |
| |
| -- verify that this doesn't merge for bucketed tables |
| create table foo_n6 (a bigint, b string) clustered by (a) into 256 buckets; |
| create table bar_n1 (a bigint, b string); |
| insert overwrite table foo_n6 select * from bar_n1; |
| dfs -ls ${hiveconf:hive.metastore.warehouse.dir}/foo_n6; |
