[ 
https://issues.apache.org/jira/browse/PHOENIX-4195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16508977#comment-16508977
 ] 

Geoffrey Jacoby commented on PHOENIX-4195:
------------------------------------------

As [~jamestaylor] suggests, I'll move the visibility issues to a separate JIRA. 

[~tdsilva], in the particular use case in which I encountered this bug, yes, I 
can use a tenant connection and delete from the view directly. (Actually, we 
were already supposed to be but weren't.) 

That's not really the point though. The Phoenix/HBase combo is designed to be 
used as a SQL-based system of record. That means that if you say DELETE FROM 
Foo Where <predicate>, and if the query returns successfully then all the rows 
for which that predicate is true should be deleted. That's not a nice-to-have, 
it's bare minimum functionality. 

If it throws an error in this case, as James suggests, that solves the 
correctness problem, but introduces a usability one: it means that if a single 
tenant creates a view, anyone using a global connection suddenly loses the 
ability to delete from the table. At the least that's non-intuitive. 

PHOENIX-1499, to update the view indexes, is also a problem that needs to be 
solved. But for this one, I think either the optimization needs to go away 
until the visibility issues are addressed, or for it to be flagged. I'd prefer 
the flag to default to "off", but if the consensus is "on" I will implement it 
like that and just turn it off in my own use case. 

> PHOENIX-4195 Deleting view rows with extended PKs through the base table 
> silently fails
> ---------------------------------------------------------------------------------------
>
>                 Key: PHOENIX-4195
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-4195
>             Project: Phoenix
>          Issue Type: Bug
>            Reporter: Thomas D'Silva
>            Assignee: Geoffrey Jacoby
>            Priority: Major
>         Attachments: test.diff
>
>
> The attached test fails.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to