[
https://issues.apache.org/jira/browse/GEODE-5596?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16594079#comment-16594079
]
ASF subversion and git services commented on GEODE-5596:
--------------------------------------------------------
Commit 47dc37b4b318b050c3e1356c947d10a8f582763d in geode's branch
refs/heads/feature/GEODE-5596 from [~bschuchardt]
[ https://gitbox.apache.org/repos/asf?p=geode.git;h=47dc37b ]
Merge branch 'feature/GEODE-5596' of ssh://github.com/apache/geode into
feature/GEODE-5596
> Client ends up with destroyed entry after invalidate()
> ------------------------------------------------------
>
> Key: GEODE-5596
> URL: https://issues.apache.org/jira/browse/GEODE-5596
> Project: Geode
> Issue Type: Improvement
> Components: regions
> Reporter: Bruce Schuchardt
> Assignee: Bruce Schuchardt
> Priority: Major
> Labels: pull-request-available
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> If a client is doing a destroy, putAll and invalidate in separate threads and
> the events align like this in the server: destroy then create then
> invalidate, the client may end up having a destroyed entry instead of an
> invalid entry.
> This is a long standing issue with Geode (and formerly with GemFire) that a
> client does not create an Invalid entry in some situations. A lot of this
> was cleared up when we implemented concurrency controls in the cache but this
> one corner case remains.
> Fixing this issue requires a small change in behavior for client caches in
> how they handle the invalidate() operation, but I think it's a change for the
> better.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
