Do you have a Navigator role? If so, delete that as well. ... View more

Also try to remove roles for which you don't have licenses, like reports manager. ... View more

From Serega in scm-users (consolidating discussion here): Hi, unfortunately I talk about other cluster which is not under Cloudera subscription support, so I can't create support ticket for it. It's a rule violation. Cluster with support doesn't suffer from such issue and ofcourse it has a license. I've seen a bug where right after uploading a license, you may hit this error unless you restart CM. You may also be able to just log out and log back in. As mark asked, did you start a trial license? Does restarting CM help? Are you modifying any other properties along with the mail server password? ... View more

CM doesn't manage hiveserver1 because it should basically never be used. It notably has no real support for concurrent operations, and you may get corruption in such scenarios.   Since HS2 is not set up by default, there's no way to use O/JDBC to connect to hive in CDH4 by default. You can, however, tell CM to initially install the HS2 role if you click "Inspect Roles" (or something similar) in the setup wizard, on the page where you select which service types you want to install. And as you know, you can also just add the role later. ... View more

Hi Mark,   HiveServer2 is only available in CDH4.2+, and is considered generally optional in CDH4.x. Many customers use the old Hive CLI on CDH4, since HiveServer2 is somewhat new. CM generally tries not to install optional roles, just required ones.   In CDH5, CM treats HiveServer2 is a required role for Hive, since the old Hive CLI will be deprecated in CDH5 and all clients should instead go through the beeline CLI and/or drivers that talk to HS2.   Hope this clears it up for you!   -Darren ... View more

There isn't a great way to fix this right now. We're looking to fix this for CM 5.   You can edit /usr/lib64/cmf/service/hdfs/hdfs.sh to change your ulimits, but know that this code does not run as root. I think processes will by default get the limits for the root user instead of the ones for the appropriate user, so you can also try changing the ulimits for root.   This issue isn't seen in many clusters, so it could also have something to do with your configuration, how you wrote the map reduce job, or maybe having a lot of small HDFS files. Could also be worth looking into, if only to optimize.   You can find a little more discussion on the topic here: https://groups.google.com/a/cloudera.org/forum/#!topic/scm-users/G1pRUfYNOWs   Thanks, Darren ... View more

Glad you got it working! ... View more