Your queries are not appearing on the Queries page
Check the role that you have been assigned. If you are not an administrator, then you can only view the queries that you have run. Only administrators can view queries and data of other users.
-
To enlist yourself as the admin user, go to
, and add your user name in the admin_users field. For example,hive,admin1,admin2
. - Check the configuration settings for Hive and Tez in the Ambari UI. See Check Cluster Configuration for Hive and Tez.
-
Verify whether the logs are added to the paths specified in the
hive.hook.proto.base-directory
and thetez.history.logging.proto-base-dir
directory configurations.Use the following commands to get the paths to the logs for Hive and Tez. If you have not run any queries, then the log directories may be empty.
For Hive:
hdfs dfs -ls /warehouse/tablespace/external/hive/sys.db/query_data
For Tez:
hdfs dfs -ls /warehouse/tablespace/external/hive/sys.db/dag_data
hdfs dfs -ls /warehouse/tablespace/external/hive/sys.db/dag_meta
hdfs dfs -ls /warehouse/tablespace/external/hive/sys.db/app_data
-
Verify whether the following proto directory configurations are present in the /etc/das/conf/das-event-processor.json file:
hive.hook.proto.base-directory
andtez.history.logging.proto-base-dir
. If these configurations are not present in the /etc/das/conf/das-event-processor.json file, then do not edit thedas-event-processor.json
file - because they will be overwritten by the configurations specified in Ambari.At this stage, restart the DAS service and recheck. If this doesn't fix the problem, then you might be using multiple configuration groups in Hive and Tez.
If you are using different or multiple configuration groups in Hive or Tez, then make sure that you specify the same paths in all the configuration groups. Presently, DAS does not support multiple configuration groups.
- Verify whether you have the required access and permission to directories specified in the
hive.hook.proto.base-directory
andtez.history.logging.proto-base-dir
directory configurations by checking errors from EventProcessorPipeline in the/var/log/das/event-processor.log
file. - Make sure that the DAS services are running as the hive service user.
-
If you have a large number of users running Hive queries simultaneously using DAS, the connections from Hive can get exhausted. This sends the query to a queue in Hive which takes a long time to start running. This can result in the following:
- If query is fired from DAS, it does not return any log or indicate that the query has been queued.
- The query does not appear on the UI, because no event is received by DAS.
Evaluate your cluster usage. You may need to add more machines to the clusters to resolve this issue.