[
https://issues.apache.org/jira/browse/HDFS-3153?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13270571#comment-13270571
]
Aaron T. Myers commented on HDFS-3153:
--------------------------------------
bq. A configuration change will affect users - The URI is derived from this
configuration parameter.
The URI is already derived from a configuration parameter, it's just that that
configuration parameter is overloaded to be both the nameservice ID and the URI
which users refer to from clients. As I understand the proposal, it's not
obvious to me that this change would necessarily affect users, since the
publicly-visible URIs could theoretically remain the same both before and after
this change - only the underlying configs would change.
bq. Do you feel it is acceptable to make this change after 2.0 is shipped?
I think we should see a concrete proposal or patch to be able to make the
determination as to whether or not 2.0.0 can ship without this change. Only
then can we determine whether or not this change can be made in a
backward-compatible fashion. Regardless of that, however, I agree with Todd and
Eli that it seems fine to ship 2.0.0-alpha without this change, as it is a
clearly-labeled alpha release.
> For HA, a logical name is visible in URIs - add an explicit logical name
> ------------------------------------------------------------------------
>
> Key: HDFS-3153
> URL: https://issues.apache.org/jira/browse/HDFS-3153
> Project: Hadoop HDFS
> Issue Type: Improvement
> Affects Versions: 2.0.0
> Reporter: Sanjay Radia
> Priority: Blocker
> Fix For: 2.0.0
>
>
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
