Hi, In my env(HDP3.1 with Hive3, the same as your env), I test hive -e "use default; set mapreduce.job.reduces=1;", and it didn’t throw any error. And maybe you can modify the configuration of your Hive conf. And if you cannot fix it at Hive side, you can just use “kylin.source.hive.redistribute-flat-table” to skip step2, I think it should work(but not a good practice).
---------------- Best wishes, Xiaoxiang Yu 发件人: 冯腾 <f_t8...@163.com> 答复: "user@kylin.apache.org" <user@kylin.apache.org> 日期: 2019年10月29日 星期二 14:26 收件人: "user@kylin.apache.org" <user@kylin.apache.org> 主题: kylin#2 Step Error Error # 2 Step executed in kylin, as indicated in the log.Step by step, finally found in the hive perform “set mapreduce.job.reduces=1”.Error: Error while processing the statement: always modify graphs. The job, reduces the at runtime. It is not in the list of params that are allowed to be modified at runtime (state = 42000, code = 1) who encountered this problem.Help me out if you can
