]
Dan Berindei updated ISPN-6100:
-------------------------------
Status: Pull Request Sent (was: Open)
Git Pull Request:
L1 entries not always stored as L1InternalCacheEntry
----------------------------------------------------
Key: ISPN-6100
URL:
https://issues.jboss.org/browse/ISPN-6100
Project: Infinispan
Issue Type: Bug
Components: Core, Test Suite - Core
Affects Versions: 8.1.0.Final
Reporter: Dan Berindei
Assignee: Dan Berindei
Labels: testsuite_stability
Fix For: 8.2.0.Beta1
{{DistributedClusteringLogic.commitSingleEntry}} is using regular {{Metadata}} instead of
{{L1Metadata}} when storing entries that are not owned by the local node. The entry will
then be stored as a regular {{MortalCacheEntry}} instead of an {{L1InternalCacheEntry}}.
In non-transactional mode, I believe this can only happen if the node was an owner at the
time of entry wrapping. I'm not sure if the same applies in transactional caches.
{{BaseDistFunctionalTest.assertOwnershipAndNonOwnership()}} assumes L1 entries are stored
as {{L1InternalCacheEntries}}, so at the very least the mismatch results in random test
failures in {{BaseDistFunctionalTest}} subclasses like {{ConcurrentJoinTest}}. It may
result in stale data as well.