[
https://issues.apache.org/jira/browse/PHOENIX-7243?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17823518#comment-17823518
]
ASF GitHub Bot commented on PHOENIX-7243:
-----------------------------------------
stoty commented on PR #1838:
URL: https://github.com/apache/phoenix/pull/1838#issuecomment-1978334436
I am not against this change, I just want to understand the problem better.
Is there any reason ever not to use a server connection (or at least use the
server connection methods to create it) on the server side ?
Maybe it would be better to make sure that CQSI always uses server
connections on the server side ?)
Would that work ?
> Add isServerConnection property to ConnectionInfo class.
> --------------------------------------------------------
>
> Key: PHOENIX-7243
> URL: https://issues.apache.org/jira/browse/PHOENIX-7243
> Project: Phoenix
> Issue Type: Sub-task
> Reporter: Rushabh Shah
> Assignee: Palash Chauhan
> Priority: Major
>
> In PhoenixDriver, we have a cache of ConnectionQueryServices which is keyed
> by ConnectionInfo object. Refer
> [here|https://github.com/apache/phoenix/blob/master/phoenix-core-client/src/main/java/org/apache/phoenix/jdbc/PhoenixDriver.java#L258-L270]
> for more details.
> Lets say if we want to create a server connection (with property
> IS_SERVER_CONNECTION set to true) and we already have a _non server_
> connection present in the cache (with the same user, principal, keytab,
> haGroup), it will return the non server connection.
> We need to add isServerConnection property to
> [ConnectionInfo|https://github.com/apache/phoenix/blob/master/phoenix-core-client/src/main/java/org/apache/phoenix/jdbc/ConnectionInfo.java#L317-L334]
> class to differentiate between server and non server connection.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
