Member since
09-29-2015
286
Posts
601
Kudos Received
60
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
11459 | 03-21-2017 07:34 PM | |
2883 | 11-16-2016 04:18 AM | |
1608 | 10-18-2016 03:57 PM | |
4266 | 09-12-2016 03:36 PM | |
6213 | 08-25-2016 09:01 PM |
03-09-2016
02:57 AM
3 Kudos
In a real cluster dfs.namenode.acls.enabled is set to true. The sandbox is configured for a wide audience to run into the least amount of problems doing the tutorials. And no, the acl command does not generate Ranger policies. You set the policies in Ranger and it will manage the ACLs for HDFS as the dfs.namenode.inode.attributes.provider.class is set to org.apache.ranger.authorization.hadoop.RangerHdfsAuthorizer See Slides 18 and 19 in http://www.slideshare.net/Hadoop_Summit/securing-hadoop-with-apache-ranger If there are no Ranger policies, it falls back to HDFS ACL
... View more
03-09-2016
12:59 AM
2 Kudos
As the zookeeper user, after a successful kinit, in a Kerberos enabled cluster,we still cannot invoke zkCli.sh. The error is: 2016-03-04 12:08:47,596 - ERROR
[main-SendThread(localhost.localdomain:2181):ZooKeeperSaslClient@384] -
An error: (java.security.PrivilegedActionException:
javax.security.sasl.SaslException: GSS initiate failed [Caused by
GSSException: No valid credentials provided (Mechanism level: Server not
found in Kerberos database (7))]) occurred when evaluating Zookeeper
Quorum Member's received SASL token. Zookeeper Client will go to
AUTH_FAILED state.
... View more
Labels:
- Labels:
-
Apache Zookeeper
-
Kerberos
-
Security
03-09-2016
12:40 AM
1 Kudo
No. For now, Voltage is supported as a KMS. The only other supported option is Ranger KMS.
... View more
03-09-2016
12:22 AM
No it can be the same metadata repository. You have two different MySQL instances for two different metadata repositories if you wish
... View more
03-08-2016
04:34 AM
2 Kudos
Yes you create another HS2 with HTTP transport to bind. Knox only uses http.
... View more
03-07-2016
07:33 PM
It has now been patched. https://issues.apache.org/jira/browse/HIVE-11470
... View more
03-04-2016
06:38 AM
2 Kudos
@Predrag Minovic
Do a vi /etc/ranger/usersync/conf/ranger-ugsync-site.xml What is in your ranger.usersync.source.impl.class property? Yes try in a chron job run for Unix: java -Dlogdir=/var/log/ranger/usersync -cp "/usr/hdp/current/ranger-usersync/dist/unixusersync-0.5.0.2.3.2.0-2950.jar:/usr/hdp/current/ranger-usersync/lib/*:/etc/ranger/usersync/conf" org.apache.ranger.unixusersync.process.UnixUserGroupBuilder OR for LDAP java -Dlogdir=/var/log/ranger/usersync -cp "/usr/hdp/current/ranger-usersync/dist/unixusersync-0.5.0.2.3.2.0-2950.jar:/usr/hdp/current/ranger-usersync/lib/*:/etc/ranger/usersync/conf" org.apache.ranger.unixusersync.process.PolicyMgrUserGroupBuilder
... View more
03-04-2016
02:51 AM
1 Kudo
@Kumar Sanyam Your DataNode cannot connect to your Name Node: So either: Your NameNode went down Your firewall prevents you from connecting from Data Node to Name Node your /etc/hosts is wrong. Remove any reference to 127.0.0.1. Ensure name Node FDQN is there. Do a netstat and ensure that Name Noe is listening to right port. The fs.default.name property in core-site.xml doesn't point to the correct hostname for the Name Node
... View more
03-03-2016
11:13 PM
2 Kudos
Did you upgrade Ambar first to 2.2.1?
... View more
03-03-2016
02:17 AM
1 Kudo
Did you do an upgrade? This usually occurs after an upgrade of Ambari. Did you remember to upgrade the Ambari database schema?
This is an indication that the configurations e.g. Hive configurations in the newer version was not upgraded properly.
The easiest way to fix is to look at each component, especially YARN or Hive, see if there is a missing property, usually on the initial Settings table (e.g, Hive authentication) and add it manually to the configuration file e.g. hive-site.xml.
... View more