扫二维码与项目经理沟通
我们在微信上24小时期待你的声音
解答本文疑问/技术咨询/运营咨询/技术建议/互联网交流
这篇文章主要为大家展示了“hive脚本的示例分析”,内容简而易懂,条理清晰,希望能够帮助大家解决疑惑,下面让小编带领大家一起研究并学习一下“hive脚本的示例分析”这篇文章吧。
寻甸网站制作公司哪家好,找创新互联!从网页设计、网站建设、微信开发、APP开发、响应式网站设计等网站项目制作,到程序开发,运营维护。创新互联于2013年成立到现在10年的时间,我们拥有了丰富的建站经验和运维经验,来保证我们的工作的顺利进行。专注于网站建设就选创新互联。
#!/bin/sh
date_yes=$1
if [ ! ${date_yes} ]; then
date_yes=`date -d "-1 days" +"%Y%m%d"`
fi
echo ${date_yes}
date_bef_yes=`date -d "${date_yes} -1 day" +"%Y%m%d"`
date_bef_7=`date -d "${date_yes} -7 day" +"%Y%m%d"`
echo ${date_bef_yes} ${date_bef_7}
remote_host="hadoop@192.168.1.27"
sour_addr="/home/data/superhero/"
local_addr="/data/superhero"
data_base="superhero_bi"
ser="\\\", \\\""
map1="\\\", \\\""
map2="\\\": \\\""
ios="ios"
pub="pub"
rsync -arvzP ${remote_host}:${sour_addr}/viso_config/*_${date_yes} /data/superhero/bi_tool/viso_config/ &&\
#rsync --progress -arvzP ${remote_host}:${sour_addr}/reg_act/act_${date_yes} ${local_addr}/act/act_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/reg_act/reg_${date_yes} ${local_addr}/reg/reg_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/info_${date_yes} ${local_addr}/info/info_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/all_pet_${date_yes} ${local_addr}/pet/all_pet_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/equip_${date_yes} ${local_addr}/equip/equip_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/item_${date_yes} ${local_addr}/item/item_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/card_${date_yes} ${local_addr}/card/card_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/card_super_step_${date_yes} ${local_addr}/super_step/card_super_step_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/all_scores_${date_yes} ${local_addr}/scores/all_scores_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/paylog/paylog_${date_yes} ${local_addr}/paylog/paylog_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/action_log/action_log_${date_yes} ${local_addr}/action_log/action_log_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/reg_act/mix_hc_${date_yes} ${local_addr}/mix_hc/mix_hc_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/reg_act/ios_hc_${date_yes} ${local_addr}/ios_hc/ios_hc_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/log_redis/vip_info_${date_yes} ${local_addr}/vip_info/vip_info_${date_yes} && \
#rsync --progress -arvzP ${remote_host}:${sour_addr}/spendlog/spendlog_${date_yes} ${local_addr}/spendlog/spendlog_${date_yes} && \
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/paylog/paylog_${date_yes}' overwrite into table ${data_base}.raw_paylog partition (ds='${date_yes}');
insert overwrite table mid_paylog_all_ext partition(ds='${date_bef_yes}') select order_id,admin,gift_coin,level,old_coin,order_coin,order_money,order_time,platform_2,product_id,raw_data,reason,scheme_id,user_id from mid_paylog_all where ds='${date_bef_yes}';
insert overwrite table mid_paylog_all_ext partition(ds='${date_yes}') select order_id,admin,gift_coin,level,old_coin,order_coin,order_money,order_time,platform_2,product_id,raw_data,reason,scheme_id,user_id from raw_paylog where ds='${date_yes}';
insert overwrite table mid_paylog_all partition(ds='${date_yes}') select order_id,admin,gift_coin,level,old_coin,order_coin,order_money,order_time,platform_2,product_id,raw_data,reason,scheme_id,user_id from mid_paylog_all_ext;
insert overwrite table mid_gs_user partition(ds='${date_yes}') select distinct user_id from mid_paylog_all where ds='${date_yes}' and lower(platform_2) = 'admin_test';
alter table mid_paylog_all_ext drop partition(ds='${date_bef_yes}');
alter table mid_paylog_all_ext drop partition(ds='${date_yes}');
alter table mid_paylog_all drop partition(ds='${date_bef_7}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/act/act_${date_yes}' overwrite into table ${data_base}.raw_act partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/reg/reg_${date_yes}' overwrite into table ${data_base}.raw_reg partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/info/info_${date_yes}' overwrite into table ${data_base}.raw_info partition (ds='${date_yes}');
insert overwrite table mid_info_all_ext partition (ds='${date_bef_yes}') select uid,account,nick,platform_2,device,create_time,fresh_time,vip_level,level,zhandouli,food,metal,energy,nengjing,zuanshi,qiangnengzhichen,chaonengzhichen,xingdongli,xingling,jinbi,lianjingshi,shenen,gaojishenen,gaojinengjing,jingjichangdianshu from mid_info_all where ds='${date_bef_yes}';
insert overwrite table mid_info_all_ext partition (ds='${date_yes}') select uid,account,nick,platform_2,device,create_time,fresh_time,vip_level,level,zhandouli,food,metal,energy,nengjing,zuanshi,qiangnengzhichen,chaonengzhichen,xingdongli,xingling,jinbi,lianjingshi,shenen,gaojishenen,gaojinengjing,jingjichangdianshu from raw_info where ds='${date_yes}';
insert overwrite table mid_info_all partition(ds='${date_yes}') select t.uid,t.account,t.nick,t.platform_2,t.device,t.create_time,t.fresh_time,t.vip_level,t.level,t.zhandouli,t.food,t.metal,t.energy,t.nengjing,t.zuanshi,t.qiangnengzhichen,t.chaonengzhichen,t.xingdongli,t.xingling,t.jinbi,t.lianjingshi,t.shenen,t.gaojishenen,t.gaojinengjing,t.jingjichangdianshu from (select *, row_number() over (distribute by uid sort by fresh_time desc ) as rn from mid_info_all_ext ) t where t.rn<2;
alter table mid_info_all_ext drop partition(ds='${date_bef_yes}');
alter table mid_info_all_ext drop partition(ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/pet/all_pet_${date_yes}' overwrite into table ${data_base}.raw_pet partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/equip/equip_${date_yes}' overwrite into table ${data_base}.raw_equip partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/item/item_${date_yes}' overwrite into table ${data_base}.raw_item partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/card/card_${date_yes}' overwrite into table ${data_base}.raw_card partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/super_step/card_super_step_${date_yes}' overwrite into table ${data_base}.raw_super_step partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/scores/all_scores_${date_yes}' overwrite into table ${data_base}.raw_scores partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/action_log/action_log_${date_yes}' overwrite into table ${data_base}.raw_action_log partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/mix_hc/mix_hc_${date_yes}' overwrite into table ${data_base}.raw_mix_hc partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/ios_hc/ios_hc_${date_yes}' overwrite into table ${data_base}.raw_ios_hc partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/vip_info/vip_info_${date_yes}' overwrite into table ${data_base}.raw_vip_info partition (ds='${date_yes}');
"
hive -S -e "
use ${data_base};
load data inpath '${local_addr}/spendlog/spendlog_${date_yes}' overwrite into table ${data_base}.raw_spendlog partition (ds='${date_yes}');
insert overwrite table mid_spendlog_no_gs partition(ds='${date_yes}') select a.order_id,a.user_id,a.level,a.subtime,a.coin_num,a.coin_1st,a.coin_2nd,a.goods_type,a.goods_subtype,a.goods_name,a.goods_num,a.goods_price,a.goods_cnname,a.args from (select * from raw_spendlog where ds='${date_yes}')a left outer join (select user_id from mid_gs_user where ds='${date_yes}')b on a.user_id=b.user_id where b.user_id is null ;
"
hive -S -e "
use ${data_base};
load data local inpath '/data/superhero/bi_tool/viso_config/ios_ser_list_${date_yes}' overwrite into table ${data_base}.raw_ser_list partition(ds='${date_yes}', plat='${ios}');
insert overwrite table mid_ser_list partition(ds='${date_yes}', plat='${ios}') select explode(split(substring(ser,3,length(ser)-4),'${ser}')) from ${data_base}.raw_ser_list where ds='${date_yes}' and plat='${ios}';
"
hive -S -e "
use ${data_base};
load data local inpath '/data/superhero/bi_tool/viso_config/ios_father_server_map_${date_yes}' overwrite into table ${data_base}.raw_ser_map partition(ds='${date_yes}',plat='${ios}');
insert overwrite table mid_ser_map partition(ds='${date_yes}',plat='${ios}') select explode(str_to_map(substring(ser,3,length(ser)-4),'${map1}','${map2}')) from ${data_base}.raw_ser_map where ds='${date_yes}' and plat='${ios}';
"
hive -S -e "
use ${data_base};
load data local inpath '/data/superhero/bi_tool/viso_config/pub_ser_list_${date_yes}' overwrite into table ${data_base}.raw_ser_list partition(ds='${date_yes}', plat='${pub}');
insert overwrite table mid_ser_list partition(ds='${date_yes}', plat='${pub}') select explode(split(substring(ser,3,length(ser)-4),'${ser}')) from ${data_base}.raw_ser_list where ds='${date_yes}' and plat='${pub}';
"
hive -S -e "
use ${data_base};
load data local inpath '/data/superhero/bi_tool/viso_config/pub_father_server_map_${date_yes}' overwrite into table ${data_base}.raw_ser_map partition(ds='${date_yes}',plat='${pub}');
insert overwrite table mid_ser_map partition(ds='${date_yes}',plat='${pub}') select explode(str_to_map(substring(ser,3,length(ser)-4),'${map1}','${map2}')) from ${data_base}.raw_ser_map where ds='${date_yes}' and plat='${pub}';
"
##-- insert overwrite table mid_gs_user partition(ds='${date_yes}') select distinct user_id from raw_paylog where ds='${date_yes}' and lower(platform_2) = 'admin_test'; -- 获取gs
##--
##-- insert overwrite table mid_spendlog_no_gs partition(ds='${date_yes}') select a.* from (select * from raw_spendlog where ds='${date_yes}')a left outer join (select user_id from mid_gs_user where ds='${date_yes}')b on a.user_id=b.user_id where b.user_id is null; -- spendlog剔除gs
##--
##-- -- 产生total_info
##-- insert overwrite table mid_info_all_ext partition (ds='${date_bef_yes}') select uid,account,nick,platform_2,device,create_time,fresh_time,vip_level,level,zhandouli,food,metal,energy,nengjing,zuanshi,qiangnengzhichen,chaonengzhichen,xingdongli,xingling,jinbi,lianjingshi,shenen,gaojishenen,gaojinengjing,jingjichangdianshu from mid_info_all where ds='${date_bef_yes}';
##-- insert overwrite table mid_info_all_ext partition (ds='${date_yes}') select uid,account,nick,platform_2,device,create_time,fresh_time,vip_level,level,zhandouli,food,metal,energy,nengjing,zuanshi,qiangnengzhichen,chaonengzhichen,xingdongli,xingling,jinbi,lianjingshi,shenen,gaojishenen,gaojinengjing,jingjichangdianshu from raw_info where ds='${date_yes}';
##--
##-- insert overwrite table mid_info_all partition(ds='${date_yes}') select t.uid,t.account,t.nick,t.platform_2,t.device,t.create_time,t.fresh_time,t.vip_level,t.level,t.zhandouli,t.food,t.metal,t.energy,t.nengjing,t.zuanshi,t.qiangnengzhichen,t.chaonengzhichen,t.xingdongli,t.xingling,t.jinbi,t.lianjingshi,t.shenen,t.gaojishenen,t.gaojinengjing,t.jingjichangdianshu from (select *, row_number() over (distribute by uid sort by fresh_time desc ) as rn from mid_info_all_ext ) t where t.rn<2;
##--
##-- alter table mid_info_all_ext drop partition(ds='${date_bef_yes}');
##-- alter table mid_info_all_ext drop partition(ds='${date_yes}');
##--
##-- -- 产生活跃account
##-- insert overwrite table mid_act_account partition(ds='${date_yes}') select info_all.account, act.plat, act.platform_2,act.ser from (select *, substring(uid,1,length(uid)-7) as ser from raw_act where ds='${date_yes}') act join (select uid, account from mid_info_all where ds='${date_yes}') info_all on act.uid=info_all.uid;
##--
##-- -- 产生新增account
##-- -- 由 raw_reg匹配出account select info_all.account, reg.plat, reg.platform_2,reg.ser from (select *, substring(uid,1,length(uid)-7) as ser from raw_reg where ds='20160406')reg join (select uid, account from mid_info_all where ds='20160406')info_all on reg.uid=info_all.uid;
##--
##-- insert overwrite table mid_new_account_ext partition (ds='${date_yes}')select acct.account, acct.plat, acct.platfom_2, acct.server from (select * from mid_act_account where ds='${date_yes}')acct left join (select account from mid_info_all where ds='${date_bef_yes}') info_all on acct.account=info_all.account where info_all.account is null; -- 尚需修改参数
##-- insert overwrite table mid_new_account partition (ds='${date_yes}') select ext.account, ext.plat, ext.platform_2, ext.server from (select *, row_number() over (distribute by account sort by server) as rn from mid_new_account_ext)ext where ext.rn=1;
##-- alter table mid_new_account_ext drop partition(ds='${date_yes}');
##"
以上是“hive脚本的示例分析”这篇文章的所有内容,感谢各位的阅读!相信大家都有了一定的了解,希望分享的内容对大家有所帮助,如果还想学习更多知识,欢迎关注创新互联行业资讯频道!
我们在微信上24小时期待你的声音
解答本文疑问/技术咨询/运营咨询/技术建议/互联网交流