We have issue to related to the cache replication among the replicated cluster member if the cache entry created from secondary custom load process.  Here is our setup:

1.  Configure the replicated cache using tcp-nio protocol.

2.  Configure two persistent store: build in file store and customer loader.

3.  Using the read-though magic to retrieve the cache entry if it is not already in persistent file store.

It works fine locally and event save to the local file.

However, the newly generated cache entry will not replicated to other configured node in this cluster.  What we found is that only those data replicated in cluster is created using the cache.put method.

We would like to know if there is any special configured we need to follow to make it work, or any other trick.

Thanks

Louie