[
https://issues.apache.org/jira/browse/PHOENIX-5584?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16981294#comment-16981294
]
Chinmay Kulkarni commented on PHOENIX-5584:
-------------------------------------------
[~larsh] I guess we should also update the server-side cache for the ptable of
the view after combining all columns from its parent hierarchy, right? This is
currently done in MetaDataEndpointImpl#buildTable.
Also, can we move the connection creation try-with-resources block outside of
the while loop?
> old client can't get right view metadata when 4.15 client created view
> ----------------------------------------------------------------------
>
> Key: PHOENIX-5584
> URL: https://issues.apache.org/jira/browse/PHOENIX-5584
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.15.0, 5.1.0
> Reporter: Xinyi Yan
> Assignee: Chinmay Kulkarni
> Priority: Blocker
> Fix For: 4.15.0, 5.1.0
>
> Attachments: 5584-4.x-HBase-1.5-v2.txt, 5584-4.x-HBase-1.5.txt
>
>
> To reproduce this bug locally:
> 1. start the sandbox at 4.15.
> 2. created table and view at 4.15 client.
> {code:java}
> CREATE TABLE AAA (A BIGINT PRIMARY KEY, B BIGINT);
> CREATE VIEW A_VIEW AS SELECT * FROM AAA;
> {code}
> 3. query `SELECT * FROM A_VIEW` at 4.14 client.
> {code:java}
> 0: jdbc:phoenix:localhost:51070> SELECT * FROM A_VIEW;
> +--+
> | |
> +--+
> +--+
> No rows selected (0.009 seconds)
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
