kylin-扫描10 439 900行后协处理器超时

wj8zmpe1  于 2021-07-13  发布在  Hbase
关注(0)|答案(0)|浏览(248)

在kylin insight中执行查询时扫描10 439 900行后获取协处理器超时。
我现在的数据是每年1.5亿。
版本
Kylin-3.0.1
hadoop-2.7.0版
hbase-1.1.2版
在中的道具下方添加 kylin.properties ```
kylin.query.scan.threshold=5000000000
kylin.storage.hbase.coprocessor-mem-gb=20
kylin.storage.partition.max-scan-bytes=0

异常跟踪
执行sql时扫描10439900行后协处理器超时:“select sum(gwp)gwp,sum(nwp)nwp,sum(gwp\u od)gwp\u od,sum(gwp\u tp)gwp\u tp,sum(nwp\u od)nwp\u od,sum(discount)gwp\u od,sum(discount)gwp\u od,sum(discount)nwp\u od from(select sum(ins\u coverage\u premium)gwp,sum(ins\u nwp)nwp,sum(case when ins\u coverage\u category='od'然后ins\u coverage\u premium else 0 end)gwp\u od,sum(case when ins\u coverage\u category in('tp','pa')然后ins\u coverage\u premium else 0 end)gwp\u tp,sum(case when ins\u coverage\u category='od'然后ins\u nwp else 0 end)nwp\u od,sum(case when ins\u coverage\u category in('tp','pa')然后ins\u nwp else 0 end)nwp\u tp,sum(当ins\ U COVERSION\ U category in('tp','pa')和ins\ U COVERSION\ U type='dc'然后ins\ U COVERSION\ U premium else 0 end时的情况)折扣\U gwp\ od,sum(当ins\ U COVERSION\ U category in('tp','pa')时的情况,'pa')和ins\u coverage\u type='dc'然后ins\u nwp else 0 end)折扣\u nwp\u od来自rsa\u kpi\u fact\u ins\u new ins left join kpi\u fin year\u master as kpi\u fin year\u master on ins.financial\u year\u master.fin year\u left join kpi\u fin month\u master as kpi\u fin month\u master on ins.eff\u fin year\u month=kpi\u fin month\u master.fin\u month left joinkpi\u product\u master as kpi\u product\u master on ins.product\u code=kpi\u product\u master.product\u code left join kpi\u branch\u master as kpi\u branch\u master on ins.branch\u code=kpi\u branch\u master.branch\u code left join kpi\u campaign\u master as kpi\u campaign\u master.campaign\u code left join kpi\u oa\u master nw askpi\u oa\u master\u nw on ins.oa\u code=kpi\u oa\u master\u nw.oa\u code left join kpi\u model\u master\u nw as kpi\u model\u master\u nw on ins.make=kpi\u model\u master\u nw.make and ins.modelcode=kpi\u model\u master\u nw.model\u code left join kpi\u sub\u channel\u master\u nw as kpi\u sub\u channel\u master\u nw on ins.channel=kpi\u sub\u channel\u name and ins.sub\u channel=kpi\子\渠道\主控\ nw.sub\渠道左加入kpi\业务\类型\主控作为kpi\业务\类型\主控在ins.business\类型=kpi\业务\类型\主控。业务\类型左加入kpi\策略\类别\主控\ nw作为kpi\策略\类别\主控\ nw在ins.policy\类别=kpi\策略\类别\主控\ nw.policy\类别,其中(ins.fin\日期>='2018-04-01'和ins.fin\u date<'2020-03-31')按ins.policy\u category、ins.product\u code、ins.oa\u code、ins.campain\u code、ins.channel、ins.sub\u channel、ins.financial\u year、ins.eff\u fin\u year\u month、ins.business\u type、ins.branch\u code、ins.uw\u year、ins.modelcode、ins.make、kpi\u model\u master\u nw.modelgroup分组。“

暂无答案!

目前还没有任何答案,快来回答吧!

相关问题