[
https://issues.apache.org/jira/browse/PHOENIX-5524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16951098#comment-16951098
]
Geoffrey Jacoby commented on PHOENIX-5524:
------------------------------------------
[~larsh] - note that the current behavior in all currently released versions of
Phoenix is _no_ SCN mutations at all. PHOENIX-4983 loosened things up, but
hasn't been released yet. It relaxed a bit too much, so we want to tweak the
rule before 4983 is ever released, so we're never "taking something away".
That's why 4.15 rather than 4.15.1 is important.
[~ckulkarni] - yes, it's just adding global immutable tables to the existing
check. To Lars's point, I _think_ that locals would be OK to use an SCN on an
upsert.
> Connections with SCN should fail mutations on tables with any index
> -------------------------------------------------------------------
>
> Key: PHOENIX-5524
> URL: https://issues.apache.org/jira/browse/PHOENIX-5524
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.15.0, 5.1.0
> Reporter: Priyank Porwal
> Priority: Major
> Fix For: 4.15.0, 5.1.0
>
>
> PHOENIX-4983 allowed mutations on subset of tables for connections using SCN.
> We need to forbid mutations on tables with any index so that they do not
> conflict with assumptions made by consistent secondary indexing redesign.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
