It is seems that in case of a large put all (for instance having 10.000 entries) operation, it is possible to get a back pressure indexing exception:

ISPN014067: Hibernate Search updates are not keeping up. Look into increasing index writer queue and/or thread pool sizes. 

We want to try to avoid the chance to get this error, limiting the indexing rate flow in case of a large (more than 100 items) put all.