[
https://issues.jboss.org/browse/ISPN-6405?page=com.atlassian.jira.plugin....
]
William Burns commented on ISPN-6405:
-------------------------------------
I figured out the core issue here. The problem is that due to the in memory expiration
being asynchronous, this completes and creates a bunch of asynchronous expiration
requests. It then proceeds to process the store synchronously. It is then possible for
the store to read the entry and attempt to remove it after the asynchronous thread has
since completed (thus getting 2 notifications). The true fix is thus to make the
processExpiration for in memory entries to be synchronous. And to be honest this fix is a
lot better since before we could spin up and keep all async threads busy if there are
enough expired in memory entries which is what we don't want.
Persistence configuration with clustered cache can cause duplicate
expiration messages
--------------------------------------------------------------------------------------
Key: ISPN-6405
URL:
https://issues.jboss.org/browse/ISPN-6405
Project: Infinispan
Issue Type: Bug
Components: Expiration
Reporter: William Burns
Assignee: William Burns
It seems that some configuration of persistence causes the entry to be expired twice when
it should have only expired once.
--
This message was sent by Atlassian JIRA
(v6.4.11#64026)