[
https://issues.apache.org/jira/browse/PHOENIX-4195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16506281#comment-16506281
]
Geoffrey Jacoby commented on PHOENIX-4195:
------------------------------------------
A config flag makes me nervous.
For a flag on this optimization to prevent an unexpected extra RPC call, it
would have to default to optimization on. (If it defaulted to off, lots of
users wouldn't notice the toggle was there, and get the extra RPC call,
especially since _clients_ would have to change.)
But that goes both ways. If the toggle defaults to "on", then we're requiring
users to read release notes really closely to realize they need to turn this
new switch on to get a bug fix. Not only that, but since in many organizations
there will be two different teams – one who manages the HBase clusters and the
Phoenix settings, and one who actually writes the queries, then you're putting
the burden on infra orgs to not only know about the new flag, but determine
whether the optimization is safe. That means they need to know every table,
view, and query in their system.
In many larger organizations, that would be a tall order, and in some cloud
setups, the infra and application teams might even be in different _companies._
If we have a flag, it should default to no optimization. Correctness, even in
an edge case, is more important than preventing an extra RPC call.
> 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)
