[infinispan-issues] [JBoss JIRA] (ISPN-3348) Unable to read entries from SingleFileCacheStore after server restart

Mircea Markus (JIRA) jira-events at lists.jboss.org
Wed Jul 24 09:27:27 EDT 2013


     [ https://issues.jboss.org/browse/ISPN-3348?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mircea Markus updated ISPN-3348:
--------------------------------

    Priority: Critical  (was: Major)

    
> Unable to read entries from SingleFileCacheStore after server restart
> ---------------------------------------------------------------------
>
>                 Key: ISPN-3348
>                 URL: https://issues.jboss.org/browse/ISPN-3348
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Server
>    Affects Versions: 6.0.0.Alpha1
>            Reporter: Martin Gencur
>            Assignee: Galder Zamarreño
>            Priority: Critical
>             Fix For: 6.0.0.Alpha3
>
>
> Using the following configuration:
> {code:xml}
> <subsystem xmlns="urn:infinispan:server:core:5.3" default-cache-container="default">
>             <cache-container name="default" default-cache="default" listener-executor="infinispan-listener" eviction-executor="infinispan-eviction" replication-queue-executor="infinispan-repl-queue">
>                 <local-cache name="default" start="EAGER">
>                     <locking isolation="NONE" acquire-timeout="30000" concurrency-level="1000" striping="false"/>
>                     <transaction mode="NONE"/>
>                     <store class="org.infinispan.loaders.file.SingleFileCacheStore" passivation="false" preload="false" purge="false">
>                         <property name="location">${java.io.tmpdir}/single-file-cache-store</property>
>                         <property name="maxEntries">2</property>
>                     </store>
>                 </local-cache>
>             </cache-container>
>         </subsystem>
> {code}
> ...it is not possible to retrieve the cache entry from the cache store after server restart, as demonstrated by the following test:
> {code:java}
> RemoteCache<String, String> rc = rcm.getCache();
>         rc.clear();
>         assertNull(rc.get("k1"));
>         rc.put("k1", "v1");
>         rc.put("k2", "v2");
>         rc.put("k3", "v3");
>         assertEquals("v1", rc.get("k1"));
>         assertEquals("v2", rc.get("k2"));
>         assertEquals("v3", rc.get("k3"));
>         controller.kill(CONTAINER);
>         controller.start(CONTAINER);
>         assertEquals("v2", rc.get("k2"));
> //^^^fails here - unable to find k2
>         assertEquals("v3", rc.get("k3"));
>         assertNull(rc.get("k1")); //maxEntries was 2, this entry should be lost as the oldest entries are removed
>         controller.stop(CONTAINER);
> {code}
> Note that preload is set to false for this test. When I set it to true, the test passes.
> I tried to modify BaseCacheStoreFunctionalTest#testPreloadAndExpiry, disable "preload" and retrieve the cache entries after server restart by calling cache.get() (not from DataContainer) and the test passed (I ran SingleFileCacheStoreFunctionalTest which extends BaseCacheStoreFunctionalTest). So it seems the bug is related to the server distribution.

--
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