It controls the ulimit for maximum file descriptors for the given process. There's one version of this property for each daemon role. The operating system enforces that each process doesn't have more open files than the configured limit. ... View more

Hi Elad, You can set probably dfs.datanode.max.locked.memory to the value you want to enforced by the OS, then use the Advanced Configuration Snippet for hdfs-site.xml to change the value that HDFS actually sees for this parameter in hdfs-site. If CM didn't set memlock, then HDFS caching would be at the mercy of the default memlock setting and could fail unexpectedly, with essentially no workaround. So it's pretty important that CM sets this limit to something that will make caching work. Thanks, Darren ... View more

Hi Bart, There's not really a problem here. Hive emits that warning, but we've found that in some cases setting metastore.local will avoid bugs (specifically when running on postgresql). So better to have the warning than to have the bug. This might help explain the process directory where you're finding all of those hive-site.xml files: http://blog.cloudera.com/blog/2013/07/how-does-cloudera-manager-work/ CM will automatically generate the metastore uris property based on the configured Hive Metastore Server role's host and port. Unless you have "Bypass Hive Metastore Server" selected, CM will emit the metastore uris property in all client configuration (/etc/hive/conf/hive-site.xml) and all dependent services (impala, hue, etc) that need to talk to Hive. The hive deprecation warning is thrown whether or not the metastore uris are configured. So you should just ignore this error. Thanks, Darren ... View more