Enable paging when retrieving data

…to avoid to put too much memory pressure on the Cassandra nodes. This should solve the problem which was leading to Cassandra nodes crash (Out of Memory) when a user was requesting a huge amount of data and should trigger garbage collections more often on the Cassandra nodes but much shorter GC stop-of-the-world pauses on the Cassandra nodes than what we observed before. With this change, the Cassandra timeout can be lower down to a more reasonable value (a few seconds).