[infinispan-issues] [JBoss JIRA] (ISPN-2939) Remove option for async Marshalling
Michal Linhard (JIRA)
jira-events at lists.jboss.org
Tue Mar 19 11:05:41 EDT 2013
[ https://issues.jboss.org/browse/ISPN-2939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12761846#comment-12761846 ]
Michal Linhard commented on ISPN-2939:
--------------------------------------
Per Sanne's request, I'm appending an example of an _expected_ inconsistency, that can occur with async marshalling:
let's say we have cache c1 configured to be DIST_ASYNC with numowners=2 and asyncMarshalling=true, running on 2 nodes (in second node let's call it c2)
let's do following 2 put operations and follow the value of key "A" in the cache on the two nodes:
{code}
c1 c2
------------------------
null null
c1.put("A", new BigValue()); // new BigValue()takes much longer to marshall than "SmallValue"
BigValue null // (BigValue not replicated yet, still marshalling)
c1.put("A", "SmallValue");
SmallValue null // SmallValue not replicated yet
SmallValue SmallValue // SmallValue replicated before BigValue marshalled
SmallValue BigValue // BigValue marshalled and replicated
{code}
we have inconsistency between nodes without any error reported, all messages nicely processed.
> Remove option for async Marshalling
> -----------------------------------
>
> Key: ISPN-2939
> URL: https://issues.jboss.org/browse/ISPN-2939
> Project: Infinispan
> Issue Type: Task
> Components: Core API
> Reporter: Sanne Grinovero
> Assignee: Mircea Markus
> Fix For: 6.0.0.Beta1
>
>
> Seems to introduce inconsistencies - and generally hard to figure out how to configure properly.
--
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