Affects Version/s: None
Fix Version/s: None
Carl Steinbach made changes -
|Status||Open [ 1 ]||Resolved [ 5 ]|
|Resolution||Incomplete [ 4 ]|
YoungYik made changes -
|Field||Original Value||New Value|
|Summary||set hive.exec.reducers.max=<number> is not working to view||test|
|Issue Type||Bug [ 1 ]||Test [ 6 ]|
|Affects Version/s||0.6.0 [ 12314524 ]|
|Environment||Linux 2.6.18-128.el5 x86_64 GNU/Linux, JRE 1.6.0_14|
|Priority||Major [ 3 ]||Trivial [ 5 ]|
I create two views from one table of different partitions, and join these two views in my query, it just use 1 reducer and the tasks stay in 82% for a long time, and then failed.
So, I use set hive.exec.reducers.max=28 before the execution of query, when select on joined tables, it works, but still keep saying "Number of reduce tasks determined at compile time: 1" when on the two joined views:
1)hive -e "create view view_1(uname, login) as select uname,'this30d' from userlist where domain='$domain' and year=$YYYY and month=$MM and day=$DD and type='all' and period_days=30;"
2)hive -e "create view view_0(uname, login) as select uname,'last30d' from userlist where domain='$domain' and year=$YYYY and month=$MM and day=$DD and type='all' and period_days=30;"
3)hive -e "set mapred.reduce.tasks=28; set; select v0.login,v1.login,count(*) from view_0 v0 full outer join view_1 v1 group by v0.login,v1.login; set;"
then the output:
Total MapReduce jobs = 2
Launching Job 1 out of 2
Number of reduce tasks determined at compile time: 1
In order to change the average load for a reducer (in bytes):
In order to limit the maximum number of reducers:
In order to set a constant number of reducers:
and at last, it just still keep using 1 reducer in this query.
but if I use "set;" to print the environment, it tells me:
|Component/s||CLI [ 12313604 ]|