[
https://issues.apache.org/jira/browse/PHOENIX-2525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15068832#comment-15068832
]
Thomas D'Silva commented on PHOENIX-2525:
-----------------------------------------
The tests fail without those two changes.
The commit is required because we only resolve a table once during a
transaction. If we commit it forces the table to be re-resolved.
At the end of MetaDataClient.updateStatisticsInternal we call
connection.getQueryServices().clearTableFromCache() , is this supposed to force
the table to be re-resolved the next time?
The clientTimeStamp to HConstants.LATEST_TIMESTAMP is required because we write
the stats at the latest server timestamp. When we resolve a transactional
table, the clientTimeStamp is set to the read pointer (which can be before the
server timestamp) so it won't see the latest stats.
> Views on transactional tables return no rows
> --------------------------------------------
>
> Key: PHOENIX-2525
> URL: https://issues.apache.org/jira/browse/PHOENIX-2525
> Project: Phoenix
> Issue Type: Bug
> Reporter: Mujtaba Chohan
> Assignee: Thomas D'Silva
> Attachments: PHOENIX-2525-v2.patch, PHOENIX-2525-wip.patch,
> PHOENIX-2525.patch
>
>
> Aggregate or non-aggregate queries do not return any rows to the client for
> views created over transactional table.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
