Hello,I'm working on setting up LLAP on an Apache Hive 2.2.0/HDFS 2.7.2 cluster. I seem to have it working, but I noticed that some of the longer running tasks are getting killed after ~60s. I did see this in one of the attempt logs:
2017-08-09 20:56:58,091 [INFO] [LlapScheduler] |tezplugins.LlapTaskSchedulerService|: Scheduled timeout monitor task to run after 60000 ms
2017-08-09 20:56:58,091 [INFO] [LlapScheduler] |tezplugins.LlapTaskSchedulerService|: Preempting for task=attempt_1502310325162_0009_1_00_000000_0 on any available host
I was wondering how/if one can set this timeout to a higher value? If so, is this the value that needs to be configured?
I'm using slider 0.92, is it possible to set these via slider? if so, can anyone provide some guidance on how to do this?. I might be fishing in the wrong area here....
Also, is there a general mailing list that I could look at? or submit this message to?