1.Trash is configured by two properties, fs.trash.checkpoint.interval
and fs.trash.interval
in core-site.xml.
2.In order to use the Ambari Files View, you need to
setup an HDFS proxy user for the Ambari daemon
account.
For example, if the AmbariServer daemon is running as
root, you configure a proxy user for root
in the Ambari Web UI by selecting:
Services > HDFS > Configs > Advanced > Custom
coresite > Add Property
Add the properties and values(In core-site.xml)
hadoop.proxyuser.root.groups=* and
hadoop.proxyuser.root.hosts=*
to the file as shown above. Save the changes and restart
any affected services as indicated in the Ambari Web UI. This assumes that
Ambari was installed and runs as the root user account. If this is not the
case,
change root to the appropriate user account.
3. net.topology.script.file.name property Configuring rack awareness
/etc/hadoop/conf/topology_script.py. This script references the mapping
file
/etc/hadoop/conf/topology_mapping.data
4. In Ambari when using existing MIT KDC /active
directory to map users and groups in unix to Hadoop it uses a rule-based system
to create mappings between service principals and their related UNIX usernames.
The rules are specified using the configuration property hadoop.security.auth_to_local
=default
No comments:
Post a Comment