Commit e72576d0 by mengdongxing

Add new file

parent d681b333
set hive.skewjoin.key=100000;
set hive.optimize.skewjoin=true;
set hive.input.format=org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
set hive.optimize.bucketmapjoin = true;
set hive.optimize.bucketmapjoin.sortedmerge = true;
insert overwrite table tkdm.tkdm_base_device_pay_info partition(ds='2017-02-23')
select /*+mapjoin(a)*/ nvl(b.appid,a.appid) as appid,
nvl(b.deviceid,a.deviceid) as deviceid,
nvl(b.first_payment_ds,'2017-02-23') as first_payment_ds,
if(isnull(a.deviceid),'2017-02-23',nvl(b.last_payment_ds,'1970-01-01')) as last_payment_ds,
nvl(b.num_payment_cnt,0)+nvl(a.num_payment_cnt,0) as num_payment_cnt,
nvl(b.num_payment_day,0)+nvl(a.num_payment_day,0) as num_payment_day,
round(nvl(b.amt_payment,0) + nvl(a.amt_payment,0),2) as amt_payment
from
(select appid,
deviceid,
round(sum(amount),2) as amt_payment,
count(1) as num_payment_cnt,
1 as num_payment_day
from tkio.tkio_mid_pay_day
where ds='2017-02-23' and
deviceid!='00000000-0000-0000-0000-000000000000'
and deviceid!='' and deviceid!='0'
group by appid,deviceid
distribute by appid ,deviceid
) a
full join
(select appid,
deviceid,
first_payment_ds,
last_payment_ds,
num_payment_cnt,
num_payment_day,
amt_payment
from tkdm.tkdm_base_device_pay_info
where ds= date_sub('2017-02-23',1)
) b on a.appid=b.appid and a.deviceid=b.deviceid
use tkdm;
create EXTERNAL table tkdm_base_device_pay_info(
appid string,
deviceid string,
first_payment_ds string,
last_payment_ds string,
num_payment_cnt int,
num_payment_day int,
amt_payment double
)
PARTITIONED BY (ds string)
clustered by (appid,deviceid) sorted by(appid,deviceid) into 32 buckets
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '\t'
STORED AS ORC
location 's3://reyuntkio/warehouse/tkio/tkdm.db/tkdm_base_device_pay_info'
;
\ No newline at end of file
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment