Solved: Re: Avoid Oozie running Hive action with user cred... - Cloudera Community - 128029

Support Questions

Find answers, ask questions, and share your expertise

Avoid Oozie running Hive action with user credential

avatar
Contributor

I want to run a very simple Oozie workflow with a Hive action on a kerberized cluster.

The problem is that Hive is using my credential and not the Hive-user as it is doing through Hive View. If I change my access in Ranger for "/apps/..." then the Oozie workflow is working fine. But we don't want personal account to have access for "/apps/..." folder

How is it possible to achieve do a Hive action where don't have access to "/apps"..." folder on HDFS?

== WORKFLOW.XML ==

<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<workflow-app xmlns="uri:oozie:workflow:0.5" name="oozie_hive_kerberos_test"> 
<credentials> 
	<credential name="hcat" type="hcat"> 
	<property> 
		<name>hcat.metastore.principal</name> 
		<value>hive/_HOST@<host>.com</value> 
	</property> 
	<property> 
		<name>hcat.metastore.uri</name> 
		<value>thrift://<host>.com:9083</value> 
	</property> 
	</credential> 
</credentials> 
<start to="hive"/> 
<action cred="hcat" name="hive"> 
	<hive xmlns="uri:oozie:hive-action:0.6"> 
	<job-tracker>${resourceManager}</job-tracker> 
	<name-node>${nameNode}</name-node> 
	<query> 
		use XXXXX; 
		drop table if exists YYYY.ZZZZ; 
	</query> 
	</hive> 
	<ok to="end"/> 
	<error to="kill"/> 
	</action> 
	<kill name="kill"> 
<message>${wf:errorMessage(wf:lastErrorNode())}</message> 
</kill> 
<end name="end"/> 
</workflow-app> 

== ERROR MESSAGE ==

SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory] 
Logging initialized using configuration in /data/hadoop/yarn/local/usercache/MY_USER_NAME/appcache/application_1487006380071_0351/container_e94_1487006380071_0351_01_000002/hive-log4j.properties 
FAILED: SemanticException MetaException(message:org.apache.hadoop.security.AccessControlException: Permission denied: user=MY_USER_NAME, access=EXECUTE, inode="/apps/hive/warehouse/DATABASE.db":hdfs:hdfs:d--------- 
at org.apache.hadoop.hdfs.server.namenode.FSPermissionChecker.check(FSPermissionChecker.java:319) 
at org.apache.hadoop.hdfs.server.namenode.FSPermissionChecker.checkTraverse(FSPermissionChecker.java:259) 
at org.apache.hadoop.hdfs.server.namenode.FSPermissionChecker.checkPermission(FSPermissionChecker.java:205) 
at org.apache.ranger.authorization.hadoop.RangerHdfsAuthorizer$RangerAccessControlEnforcer.checkPermission(RangerHdfsAuthorizer.java:307) 
at org.apache.hadoop.hdfs.server.namenode.FSPermissionChecker.checkPermission(FSPermissionChecker.java:190) 
at org.apache.hadoop.hdfs.server.namenode.FSDirectory.checkPermission(FSDirectory.java:1827) 
1 ACCEPTED SOLUTION

avatar
Contributor

Problem solve by using Hive2

View solution in original post

3 REPLIES 3

avatar
Master Guru

Your "credential" section looks wrong, it should be something like this:

 <property>
         <name>hcat.metastore.uri</name>
            <value>thrift://<host>:<port></value>
         </property>
         <property>
             <name>hcat.metastore.principal</name>
             <value>hive/<host>@<realm></value>
         </property>

On every node where Oozie client is installed you can find good examples for all Oozie actions including Hive action in "/usr/hdp/current/oozie-client/doc/examples". Check file called apps/hive/workflow.xml.security under "examples" and modify job.properties to provide your "realm" and other required parameters. Also, in case of hive2 action be sure to test using HS2 server running in binary transport mode. There were some bugs in http mode on kerberized cluster. This applies only to hive2 action, the hive action you are trying should work on both transport modes.

avatar
Contributor

Thx for your reply. I solve the problem by converting the Oozie script to run Hive2.

avatar
Contributor

Problem solve by using Hive2