LLAP Service sometimes may fail to startup after HDP upgrade with the following error:    Error:: Caused by: org.apache.hadoop.security.KerberosAuthException: failure to login: for principal: hive/prakash.cloudera.com@CLOUDERA.COM from keytab /grid/0/yarn/local/usercache/hive/appcache/application_**********/container_***********_01_000001/keytabs/llap0.keytab javax.security.auth.login.LoginException: Pre-authentication information was invalid (24)   To resolve this issue, run the following command and start LLAP:  yarn app -destroy llap0 If the issue persists, then do the following   Check the hive.service.keytab file is available in all the hosts, run kinit and test the connectivity. If the file is missing, recreate the keytabs on the missing hosts either through Ambari or manually (see this article for manual keytab creation). Verify the file in HDFS and copy it to the local. Perform kinit with the copied keytab. hdfs:///user/hive/.yarn/keytabs/hive/hive.service.keytab hdfs dfs -copyToLocal /user/hive/.yarn/keytabs/hive/hive.service.keytab /tmp/hive.keytab kinit -kt /tmp/hive.keytab hive/prakash.cloudera.com@CLOUDERA.COM kinit: Preauthentication failed while getting initial credentials  If you get the error, to move the keytab file from hdfs location. hdfs dfs -cp /user/hive/.yarn/keytabs/hive/hive.service.keytab /tmp/ hdfs dfs -rm /user/hive/.yarn/keytabs/hive/hive.service.keytab Start LLAP component.     ... View more

create table prakash2(id int UNIQUE DISABLE NOVALIDATE,PRIMARY KEY(id) DISABLE NOVALIDATE);   insert into table prakash2 values(1); insert into table prakash2 values(1); insert into table prakash2 values(1);   +--------------+ | prakash2.id | +--------------+ | 1 | | 1 | | 1 | +--------------+ 3 rows selected (0.149 seconds)   As per RDBMS concepts primary key itself has unique constraint but here in hive mentioning unique constraint explicitly allowing me to insert duplicate values.   How primary key and unique constraints are different HIVE apart from the traditional RDBMS systems? ... View more

CREATE TABLE Test( recordId String, Advertiser_id STRING , Advertiser_name STRING, PRIMARY KEY (recordId) DISABLE NOVALIDATE);   insert into Test values('1','prakash','vtv'); insert into Test values('1','prakash','vtv');   Test.recordid | Test.advertiser_id | Test.advertiser_name | +-----------------------------+----------------------------------+------------------------------------+ | 1 | prakash | vtv | | 1 | prakash | vtv | +-----------------------------+----------------------------------+------------------------------------+   As per RDBMS terms Primary key constraint shouldn't allow me to create duplicate record with the same value.   How to achieve this in Hive?       ... View more