I agree with you on the fact that Redis is not necessarily adapted to store information for a long time, at least not as much as a classic RDBMS. Despite that, I’ve seen many successful projects use Redis as their primary data store. In my opinion, problems only happen when storing a very large quantity of data, to the point where it cannot fit in memory; alas, Redis has no solution beyond that point. Nevertheless, I’d see it as a good fit for an event store, at least when not dealing with a very large quantity of data. At one point, sharding between Redis instances may be easy to implement, if and only if you restrict an aggregate root to one instance.
I won’t have time to implement this in the near future but I probably will in the long run, if only to see if it is feasible.
Thank you for your feedback, it is greatly appreciated,