[infinispan-issues] [JBoss JIRA] (ISPN-2149) FineGrainedAtomicMap implementation fails to consider removal of elements
Sanne Grinovero (JIRA)
jira-events at lists.jboss.org
Thu Aug 30 16:54:32 EDT 2012
[ https://issues.jboss.org/browse/ISPN-2149?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12714958#comment-12714958 ]
Sanne Grinovero commented on ISPN-2149:
---------------------------------------
I've created unit tests, with plenty of assertions.
The bad news is that each and all assertion I've added fails :-/
>From my github repo (git://github.com/Sanne/infinispan.git) please take branch ISPN-2149 as a starting point; contains three commits, the last one includes the failing tests.
> FineGrainedAtomicMap implementation fails to consider removal of elements
> -------------------------------------------------------------------------
>
> Key: ISPN-2149
> URL: https://issues.jboss.org/browse/ISPN-2149
> Project: Infinispan
> Issue Type: Bug
> Components: Core API
> Affects Versions: 5.1.5.FINAL, 5.2.0.ALPHA1
> Reporter: Sanne Grinovero
> Assignee: Vladimir Blagojevic
> Priority: Blocker
> Labels: affects_OGM
> Fix For: 5.2.0.Alpha4, 5.2.0.Final
>
>
> Most methods in the {{org.infinispan.atomic.FineGrainedAtomicHashMapProxy<K, V>}} implementation seem to ignore the use case of removing keys from the map.
> {{keySet()}}, {{values()}}, {{entrySet()}}, {{size()}}, {{isEmpty()}}
> all ignore the fact that values in the uncommitted map might have been *removed* in the current transaction, or have *intersections* with existing keys.
> Note that failing to consider intersections has it return the wrong entries as committed values overwrite the in-flight values.
> Methods like {{containsKey}} don't check if the matched key is in fact stored in the {{DeltaAwareCacheEntry}} as a deleted entry.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
More information about the infinispan-issues
mailing list