I read that "Hazelcast List cannot be scaled beyond the capacity of a single machine" (http://docs.hazelcast.org/docs/latest/manual/html/list.html).
However, I do need to use a very big list with elements distributed across cluster nodes. How can I do that using Hazelcast or some other alternative?
You can make use of the SPI in Hazelcast to build your own.
We are currently working on a design where a queue can span multiple machines. Lists are not that much used, but queues quite a lot and queues suffer from the same problem as a list that they can't grow beyond the capacity of a single machine.
Another option would be to make use of a IMap and use the index of the list as key in the map.
