Hive and STS are installed and started on the same node, in fact my cluster is a small one to test and has only one node.
But when I try to run hive queries, they are stuck in UNDEFINED state until I kill STS aplication for example.
It seems only one application can be in RUNNING state, any idea what's wrong in my config ?
Here, yarn application -list output
yarn application -list 16/08/17 12:59:57 INFO impl.TimelineClientImpl: Timeline service address: http://xxxxxxxx.xxxxx.xxxxx.xxxxx:8188/ws/v1/timeline/ 16/08/17 12:59:57 INFO client.RMProxy: Connecting to ResourceManager at xxxxxxx.xxxxxx.xxxxx.xxxx/xx.xxx.xxx.xxx:8050 Total number of applications (application-types:  and states: [SUBMITTED, ACCEPTED, RUNNING]):2 Application-Id Application-Name Application-Type User Queue State Final-State Progress Tracking-URL application_1470665309808_0005 SparkSQL::xxxxxxxxxxxxx SPARK spark default RUNNING UNDEFINED 10% http://xxxxxxxxx.xxxxxx.xxxxx.xxxxxx:4040 application_1470665309808_0006 HIVE-3621ebc5-ee9d-4113-a88b-291a4a87d5a6 TEZ hive2 default ACCEPTED UNDEFINED 0%
Thanks for help
Can you confirm your service checks work? If not, please provide your yarn-site.xml and capacity-scheduler.xml
thanks for your help. Do you know which parameter I have to check in yarn-site.xml ?