[infinispan-issues] [JBoss JIRA] (ISPN-3460) Refactor the CacheEntry hierarchy

Krzysztof Sobolewski (JIRA) issues at jboss.org
Tue Jun 7 08:13:00 EDT 2016


    [ https://issues.jboss.org/browse/ISPN-3460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13248897#comment-13248897 ] 

Krzysztof Sobolewski commented on ISPN-3460:
--------------------------------------------

FYI: one thing that breaks because of this is preload with Functional API. There is a downcast in PersistenceManagerImpl:

{code:java}
      preloadCl.process(null, new AdvancedCacheLoader.CacheLoaderTask() {
         @Override
         public void processEntry(MarshalledEntry me, AdvancedCacheLoader.TaskContext taskContext) throws InterruptedException {
            if (loadedEntries.getAndIncrement() >= maxEntries) {
               taskContext.stop();
               return;
            }
            Metadata metadata = me.getMetadata() != null ? ((InternalMetadataImpl)me.getMetadata()).actual() : null; //the downcast will go away with ISPN-3460
            // ^^^^^^^^^^^^^^^^^^^^
            preloadKey(flaggedCache, me.getKey(), me.getValue(), metadata);
         }
      }, new WithinThreadExecutor(), true, true);
{code}

With Functional API there's no InternalMetadataImpl; there's MetaParamsInternalMetadata instead.

> Refactor the CacheEntry hierarchy
> ---------------------------------
>
>                 Key: ISPN-3460
>                 URL: https://issues.jboss.org/browse/ISPN-3460
>             Project: Infinispan
>          Issue Type: Enhancement
>          Components: Core
>            Reporter: Mircea Markus
>            Assignee: William Burns
>            Priority: Minor
>
> ATM we keep the metadata information in 3 places: CacheValues, CacheEntries and Metadata. We should move everything in the Metadata and make the CacheValues and the CacheEntries reference it.
> Also look at MimeMetadataHelper in REST cache store...



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)


More information about the infinispan-issues mailing list