bytes detected at the end
of the row! Ignoring similar problems.
r7raul1...@163.com
art_tracker_id,
a1.end_user_id,
a1.cart_track_time
FROM dw.fct_traffic_navpage_path_detl a1
WHERE a1.ds = '2015-01-19'
ANDa1.cart_tracker_id > 0
AND(a1.cart_button_position IS NULL OR length(a1.cart_button_position) = 0)
AND a1.sessn_id IN ('Y49EY895
In my test enviroment . I use hive 0.14 ,hive 1.1.0 ,the result is ok.
But in production enviroment ,the result is not correct.
r7raul1...@163.com
From: Thejas Nair
Date: 2015-04-02 16:41
To: r7raul1...@163.com
CC: dev
Subject: Re: Re: hive 0.14 on some platform return some not NULL value
I download full data from hdfs. Then load data into my table. In my test
enviroment. Everything is ok.
My production is hadoop 2.3.0-cdh 5.0.2 REDHAT 5.8 java version "1.6.0_35"
r7raul1...@163.com
From: r7raul1...@163.com
Date: 2015-04-02 16:57
To: dev
Subject: Re: Re: hi
Sorry ,I check my production jdk is java version "1.7.0_45" not java
version "1.6.0_35"
r7raul1...@163.com
From: r7raul1...@163.com
Date: 2015-04-02 17:01
To: dev
Subject: Re: Re: hive 0.14 on some platform return some not NULL value as NULL
I download full data f
0.14 lost some column info. Why? BTW, My meta database schema is
hive 0.10 not update to hive 0.14.
r7raul1...@163.com
From: Thejas Nair
Date: 2015-04-02 15:29
To: dev
Subject: Re: Is it necessary to update beelinepositive q.out files?
beeline tests have been disabled for a while and and i
0.14 lost some column info. Why? BTW, My meta database schema is
hive 0.10 not update to hive 0.14.
r7raul1...@163.com
From: r7raul1...@163.com
Date: 2015-04-02 15:36
To: dev
CC: thejas.nair
Subject: Re: Re: hive 0.14 on some platform return some not NULL value as NULL
DDL is
CREATE TABLE
I use hive 0.14 to use hive 0.10 metastroe server .The problem fixed. Now hive
0.14 return correct result.
r7raul1...@163.com
From: r7raul1...@163.com
Date: 2015-04-07 10:34
To: dev
CC: thejas.nair
Subject: Re: Re: hive 0.14 on some platform return some not NULL value as NULL
I found
select userid,count(*) from u_data group by userid order by useridwill
product MRR.
I think when the result of userid,count(*) is small(one reduce can process the
result) . This query plan can optimize to MR ?
r7raul1...@163.com
string), comb_prod_id (type: bigint)
r7raul1...@163.com
10 matches
Mail list logo