Please check if your YARN container has enough resource for MR jobs. By default, if in-memory build is used, kylin_job_conf_inmem.xml will ask for 3072MB for MapReduce Job. You could reduce the number.
2017-05-22 21:59 GMT+08:00 446463...@qq.com <446463...@qq.com>: > I found that at MR log > > Could not deallocate container for task attemptId > attempt_1494561372318_21842_r_000010_3It seems memory is not enough for > Running kylin job in mine cluster?then I google for > thatyarn.nodemanager.resource.memory-mb=24576 > yarn.scheduler.maximum-allocation-mb=24576 > it's useful? > > > 446463...@qq.com > > From: 446463...@qq.com > Date: 2017-05-22 21:14 > To: dev > Subject: Container complete event for unknown container id > Hi: > when I cube today a error like that > ``` > 0 ERROR [RMCommunicator Allocator] > org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: > Container complete event for unknown container id > container_1494561372318_21813_01_018423```on Build N-Dimension Cuboid > Data Can I help me ? > > > 446463...@qq.com >
