[
https://issues.apache.org/jira/browse/HDDS-1455?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16824312#comment-16824312
]
Arpit Agarwal commented on HDDS-1455:
-------------------------------------
Thanks for filing this [~eyang]. I think both suggestions make sense.
The reason for the different prefix is that SCM and DataNodes are providing a
separate service called HDDS, which is used by OM service that provides Ozone
on top of HDDS.
But from a usability perspective it may be better to have a single prefix.
cc [~ajayydv], [~anu].
> Inconsistent naming convention with Ozone Kerberos configuration
> ----------------------------------------------------------------
>
> Key: HDDS-1455
> URL: https://issues.apache.org/jira/browse/HDDS-1455
> Project: Hadoop Distributed Data Store
> Issue Type: Sub-task
> Reporter: Eric Yang
> Priority: Major
>
> In SetupSecureOzone.md, the naming convention for keytab files are different
> from code.
> {code}
> hdds.scm.http.kerberos.keytab
> ozone.om.http.kerberos.keytab
> {code}
> In ozone-default.xml, it is looking for:
> {code}
> hdds.scm.http.kerberos.keytab
> ozone.om.http.kerberos.keytab.file
> {code}
> For the non http version of keytab, they are branded as:
> {code}
> hdds.scm.kerberos.keytab.file
> ozone.om.kerberos.keytab.file
> {code}
> It is best to shorten the name to remove .file suffix from the code to be
> consistent with Hadoop naming convention. The second nitpick is hdds and
> ozone prefix. Is there a good reason to have distinct prefix for both that
> work closely together? How about hadoop.ozone prefix? From usability point
> of view, the current prefix are very confusing.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
