Hi Ammu, beeline has a flag --verbose=[true/false] that will output more information, I use that option regularly to get back more information about stacktraces. related: https://cwiki.apache.org/confluence/display/Hive/HiveServer2+Clients#HiveServer2Clients-BeelineCommandOptions ... View more

Hi Michael, Can you check the ambari metric collector log on that machine that is not reporting ambari metrics? I believe it should be under /var/log/ambari-metrics-collector/*.log|out ... View more

Hi Nicola, I assume you are logged in into ambari as user admin. even with your fix I believe your query is being submitted as the user ambari is logged in as and then the TEZ view uses a user that does not have permission to see the job. can you try either: creating an ambari "hive" user and run the queries logged in as that user Use ldap or valid kerberos users within ambari and run the insert also, do you have impersonation/doAs enabled in ambari hive configuration? ... View more

I Prateek, I have successfully done a downgrade from 2.4 to 2,3, but to be able to do this I had to had a complete backup of the ambari database from before the upgrade. I suspect the process would be similar. the process was: stop all ambari daemons remove all ambari packages install the old version ambari packages delete the old ambari database and load from the backup start ambari server, check if its working start ambari agents If you dont have a backup of the 2.1 database then I doubt you can safely do the downgrade, there are a number of upgrade sql scripts that run when you upgrade but there are no equivalent scripts to go backwards. ... View more

Hi Sameer, If you format a namenode, you will erase all metadata on that namenode, so the namenode will no longer be aware of the data stored under the datanodes. When you format the namenode will also get a new NamespaceId. If you run hadoop fs -ls / you will get data related to the new NamespaceId, thus you will get an empty file structure. ... View more