Hello,

we currently trying to use Hibernate-Search in a clustered environment and wanted to use infinispan for replication.

Problem is:

hibernate.search.exclusive_index_use=true
hibernate.search.default.worker.execution=async
hibernate.search.default.directory_provider=infinispan
hibernate.search.default.indexBase=infinispan

Does not work as expected. I can mass index and then search and get correct results. But if I then make changes to the searched values and reserach I still get the old search results.
The configuration works as soon as I change either hibernate.search.exclusive_index_use to false or hibernate.search.default.worker.execution to sync.
This works even in the cluster so JGroupssettings are fine to.

But the problem is the same on a single node setup with infinispan!

Can you very please have a look into this? Is async workers and exclusive index use even allowed? It seems to work very well if I use filesystem provider instead of infinispan...
Our application works really well if I either go for sync workers or inexclusive index. This problem now stops me since begin of last week.
Unfortunately neither inexclusive index nor sync workers is feasable for us because both will slow down import processes which the customer is already complaining of being to slow.

Thank you for your time!

Best regards
Marco

EDIT: We are using following Versions:
<artifactId>hibernate-search-orm</artifactId>
<version>5.3.0.Final</version>
<artifactId>hibernate-search-infinispan</artifactId>
<version>5.3.0.Final</version>
<artifactId>infinispan-query</artifactId>
<version>7.2.0.Final</version>
<artifactId>hibernate-core</artifactId>
<version>4.3.8.Final</version>