I found one JIRA（https://issues.apache.org/jira/browse/HIVE-10649）,the problem described is quite similar to mine. BUT It's not as described in HIVE-10649,my cluster has four nodes and all nodes are active.
在 2019-08-05 13:15:48，"Maria" <[email protected]> 写道： Hi, all: I want to test hive LLAP performance in my cluster: Hive2.3.0,hadoop2.7.1,tez0.9.2. While tez task always fails to schedued by tez am task, after looking into tez am log, I found many suspect informations like this one : "[TaskSchedulerEventHandlerThread] |tezplugins.LlapTaskSchedulerService |: Processing de-allocate request for task=attempt_1564973659067_0002_1_00_000000_0, state=ASSIGNED, endReason=COMMUNICATION_ERROR" Does someone know what's that mean? Thanks. Maria.
