Vikram Srivastava helped me in google groups. Here is an explanation:   The alternatives priority for HDFS is by default configured lower than MapReduce, so deploying HDFS client configs only will not update what /etc/hadoop/conf points to.   I've filed an internal issue for this to warn users that they need to deploy cluster client configs rather than individual services.   Hope it would help other hadoopers 🙂   ... View more

Magic tricks with ids didn't help. They brought more problem during migration later. We did partial migration (users+their saved pig/hive/impala scripts/queries) from sqlite to postgre and the problem gone. The problem is solved for us. ... View more

Sounds like something that could be done, so I've added a JIRA to track it. https://issues.cloudera.org/browse/IMPALA-633 ... View more

Per e-mail thread discussion 1.7 JDK was manually added before cluster upgrade, and was present within custom local yum repo at the site.   CM/CDH will pick the later JDK present to use.  Per Phillp L:   Yes. We ask yum to install jdk and if yum decides that it's appropriate to install a newer version of jdk than is already installed, that is what it will do.   If you want to keep a package on a particular version, even if a newer version is available in a repo, there is a way to make yum do that:   http://blog.serverbuddies.com/how-can-i-get-yum-to-keep-package-at-a-certain-version/       ... View more