[infinispan-issues] [JBoss JIRA] (ISPN-3089) The first write to a joiner in invalidation mode can be ignored
Mircea Markus (JIRA)
jira-events at lists.jboss.org
Mon Dec 2 18:56:06 EST 2013
[ https://issues.jboss.org/browse/ISPN-3089?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mircea Markus updated ISPN-3089:
--------------------------------
Labels: nbst (was: )
> The first write to a joiner in invalidation mode can be ignored
> ---------------------------------------------------------------
>
> Key: ISPN-3089
> URL: https://issues.jboss.org/browse/ISPN-3089
> Project: Infinispan
> Issue Type: Bug
> Components: State transfer
> Affects Versions: 5.2.6.Final, 5.3.0.Beta1
> Reporter: Dan Berindei
> Assignee: Dan Berindei
> Labels: nbst
> Fix For: 5.3.0.CR2, 5.3.0.Final
>
>
> In invalidation mode we don't wait for the initial state transfer to finish, or even for the joiner to become a member in the "write" CH, before returning to the user from {{getCache()}}.
> Writes to the joiner before it becomes a member of the write CH are not committed to the local cache, because of a check in {{EntryWrappingInterceptor.shouldWrap()}}. So it's possible for the joiner to ignore the first write completely.
> If {{StateTransferConfigurationBuilder}} enabled {{awaitInitialTransfer}} by default in invalidation mode, like it does in distributed/replicated mode, this wouldn't happen.
> The fact that it will still be possible for the user to disable {{awaitInitialTransfer}} manually shouldn't be a problem, because in invalidation mode the user should expect values to be invalidated at any time. This is just about improving the default configuration.
--
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