Hi 5cai,   Please see TCLIService.FetchResults for equivalent functionality.   To retrieve operation logs, in TFetchResultsReq specify fetchType as 1.   thanks, Mohit ... View more

Hi MattPatt,   For Hive, it is the Hive Metastore server (and not Cloudera Manager) that needs access to the JDBC driver to talk to the database.   As mentioned in the documentation, JDBC driver needs to be installed on the Cloudera Manager Server host, as well as hosts to which you assign the Hive Metastore Server,  Activity Monitor, Reports Manager, Hue Server, Sentry Server, Cloudera Navigator Audit Server, and Cloudera Navigator Metadata Server roles.   See https://www.cloudera.com/documentation/enterprise/latest/topics/cm_ig_mysql.html#cmig_topic_5_5_3 for details on the location.   thanks, Mohit ... View more

Hi Dobroff,   What CDH version are you using ?   Do you see any exceptions in HiveServer2 logs ?   thanks. ... View more

Hi syamsri,   This seems like a permission issue. As indicated by the exception, please verify if appropriate write permissions exist for table directory.   Thanks. ... View more

Hi lsegrim,   In CDH 5.7.0, you could set hive.server2.logging.operation.level to VERBOSE. You can set this configuration via the HiveServer2 safety valve hive-site.xml in Cloudera Manager.    See https://cwiki.apache.org/confluence/display/Hive/Configuration+Properties#ConfigurationProperties-hive.server2.logging.operation.level   Note that  hive.server2.logging.operation.enabled should also be set to true (which  is the default via Cloudera Manager) ... View more

Hi rmiglani,   The root cause of the error is likely in HiveServer2 logs. Could you check if you see any exceptions there?   You might also want to look logs from Spark History Server: In Cloudera Manager, g oto Spark on YARN service Instance > History Server > History Service WebUI > Click on relevant job > Click on relevant Failed Job > Click on failed stages for that job and look for the "details" section.   Possible causes:  - Double-check permissions: Make sure that the client issuing the query has a home directory in HDFS. Also, make sure Spark application history directory is configured with the right permissions. See http://www.cloudera.com/documentation/enterprise/latest/topics/admin_spark_history_server.html for details. - Memory config: If you see OutOfMemory related exceptions in Spark History Server logs, you may have to increase  driver or exectors' memory. See http://www.cloudera.com/documentation/enterprise/latest/topics/admin_hos_tuning.html for tuning guidance.   thanks.     ... View more

Hi Antoo,   The INTEGER_IDX column in COLUMNS_V2 table has always been of integer datatype (as the name of the column suggests) in HMS schemas for all db types, including Postgres and MySQL.   If it is of type VARCHAR, it will change ordering of columns, as you suggest.    When you say you " changed Hive Metastore Database from Postgresql into MySQL",  how exactly did you do that ? The schema probably got corrupted in this process (wrong column name to type mapping for example). It is likely that other tables in the schema are corrupted as well. One way to check is to create a fresh metastore schema for MySQL (in a test installation, for example) and compare your current schema with it.   thanks, Mohit  ... View more

Hi G_Arti,   Could you try using  org.apache.hive.hcatalog.pig.HCatLoader() instead of  org.apache.hcatalog.pig.HCatLoader() ?   As part of HCatalog moving to the Hive project, all client facing classes were moved to from org.apache.hcatalog to   org.apache.hive.hcatalog resulting in the org.apache.hcatalog being deprecated starting CDH 5.3.   This is documented here: http://www.cloudera.com/content/cloudera/en/documentation/core/latest/topics/cdh_rn_hive_ic.html   thanks, Mohit ... View more

Hi,   Does it work without Sentry enabled ?   thanks, Mohit   ... View more

Hi,   Just to clarify, scenario where LDAP auth is used for beeline-HS2 (i.e. client) and Kerberos auth is used for HS2-HDFS (i.e. server) is supported.   HS2 cannot be configured to used both LDAP and Kerberos for client auth (i.e. scenario where one client wants LDAP auth the another client wants Kerberos auth). Such a scenario is supported by Impala, but not Hive.   From the stacktrace, there could be one of two issues:   1) Could you make sure that HMS is correctly configured ? See Hive Metastore Server Security Configuration . Also make sure hive.metastore.sasl.enabled and hive.metastore.kerberos.keytab.file are configured for HS2 as well.   2) If HMS is correctly configured, and you're still seeing this issue, you may be running into HIVE-7764. This issue has been fixed in CDH 5.2.1 and subsequent versions.   thanks, Mohit ... View more

Hi,   Yes, scenario where LDAP is used for beeline-HS2 authentication and Kerberos for HS2-HDFS  authentication  is supported.     Please take a look at HiveServer2 Security Configuration guide for related configuration parameters.   thanks, Mohit ... View more

Hi Sree,   I don't think there is a way to list only internal tables using SHOW TABLES. You'll need to use DESCRIBE EXTENDED <TABLE> and look for the "tableType" property for each table.   (Alternatively, you could use different identifiers for internal and external table names, and use SHOW TABLES <pattern>, for example: SHOW TABLES 'internal_*' to list internal ones)   thanks, Mohit     ... View more

Hi charles_tay,   This configuration is located in "Service-Wide > Advanced" section under Hive service configuration.   thanks, Mohit   ... View more

Hi guydo,   Did you try using the " Hive Auxiliary JARs Directory" configuration in CM ?   This config allows you to specify a directory containing aux jars, for both Hive CLI and HiveServer2 services.   thanks, Mohit  ... View more