Hi Vinay Actually I succeeded using my default principal & password and using the following default url: jdbc:hive2://mycluster.bdata.corp:8443/;ssl=true;transportMode=http;httpPath=/gateway/pam/hive
Still a problem remains when trying to access Hive from spark: %spark println(sc.version) val hiveContext = new org.apache.spark.sql.hive.HiveContext(sc); leads to: 1.5.2 java.io.IOException: Login failure for ta-1t77-bdata-...@bdata.corp from keytab /local/home/ta-1t77-bdata-gal/ta-1t77-bdata-gal.keytab: javax.security.auth.login.LoginException: Client not found in Kerberos database (6) at org.apache.hadoop.security.UserGroupInformation.loginUserFromKeytab(UserGroupInformation.java:962) at org.apache.spark.sql.hive.client.ClientWrapper.<init>(ClientWrapper.scala:170) at org.apache.spark.sql.hive.HiveContext.executionHive$lzycompute(HiveContext.scala:164) at org.apache.spark.sql.hive.HiveContext.executionHive(HiveContext.scala:162) Any idea why? Cheers On Tue, 16 Aug 2016 at 21:38 tog <guillaume.all...@gmail.com> wrote: > Hi Vinay > > Thanks for the quick answer. > I am a bit puzzled by your proposal. For the time being, I have been > succesful configuring Zeppelin without shiro by adding properties to the > Spark interpreter directly. > > Would it be possible to to the same with appropriate property settings for > the jdbc setting? > > One colleague also suggested to go through normal user/password but > pointing to a Knox URI instead of the normal hive one (+ some options) > > Cheers > Guillaume > > > On Tuesday, 16 August 2016, Vinay Shukla <vinayshu...@gmail.com> wrote: > >> Guillaume, >> >> Can you try this? >> >> Enable Zeppelin authentication, via Shiro config and have Kerberos setup >> for Zeppelin. Zeppelin JDBC interpreter should automatically add logged in >> user's identity to the JDBC connection after this confg. >> >> >> Note the settings below. >> >> Thanks, >> Vinay >> >> >> >> ÃIdentity Propagation with JDBC (Hive) >> 1.Enable Authentication via Shiro Config >> –zeppelin.jdbc.auth.type = KERBEROS >> –zeppelin.jdbc.principal = principal value >> –zeppelin.jdbc.keytab.location = keytab location >> •JDBC Interpreter will add end-user as >> jdbc:hive2://HiveHost:10000/default;principal=hive/_h...@my-realm.com >> ;hive.server2.proxy.user=<$USER_NAME> >> >> On Tue, Aug 16, 2016 at 9:36 AM, Guillaume Alleon < >> guillaume.all...@gmail.com> wrote: >> >> > Hi >> > >> > I am a bit confused on the way to set up Zeppelin to use hive to access >> my >> > data on a kerberized cluster. >> > I looked at the doc and there is no mention about Kerberos settings. The >> > only place where I find some traces about that was in that pull request >> > from June: https://issues.apache.org/jira/plugins/servlet/mobile# >> > issue/ZEPPELIN-1146 >> > >> > Could someone confirm the properties that need to be set-up to configure >> > the jdbc interpreter so that I can use hive on a kerberized cluster. >> > >> > Thanks >> > Guillaume >> > >> > >> > >> > Sent from a small device >> > > > -- > PGP KeyID: 2048R/EA31CFC9 subkeys.pgp.net >
