[
https://issues.apache.org/jira/browse/PHOENIX-7038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17774318#comment-17774318
]
ASF GitHub Bot commented on PHOENIX-7038:
-----------------------------------------
stoty commented on PR #1682:
URL: https://github.com/apache/phoenix/pull/1682#issuecomment-1758877367
I am in general very wary of overriding HBase properties from a Phoenix
application, as this can very easily lead to failures for properties that must
be in sync across the cluster for HBase to work.
In my ideal world we'd have some kind of whitelist for properties that are
known to be safe to override (or an inverse blacklist)
(Of course users could still wreak havoc by modifying the hbase-site.xml
used by the client)
I would tie the id propagation part to
https://issues.apache.org/jira/browse/PHOENIX-5974.
We could either correlate that single ID with the ConnectionInfo/profile
(though it requires extra bookeepeing), or just add it alongside the query id.
> Implement Phoenix Connection Profile Metrics
> --------------------------------------------
>
> Key: PHOENIX-7038
> URL: https://issues.apache.org/jira/browse/PHOENIX-7038
> Project: Phoenix
> Issue Type: New Feature
> Reporter: Mihir Monani
> Assignee: Mihir Monani
> Priority: Major
>
> With JDBC connection profile string, Clients can provide connection profile
> name. With use of Connection Profile, different kinds of clients/use-case in
> single JVM can create connection with different configuration like time-outs,
> retries, sleep between retries.
> Connection URL Example : jdbc:phoenix:localhost:2181:CONN_PROFILE_2
> Here CONN_PROFILE_2 is connection profile name.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
