If you do this, execute the future on a different dispatcher so you aren't blocking the default dispatcher. This can cause problems if you tie up the default d

If you do this, execute the future on a different dispatcher so you aren't blocking the default dispatcher. This can cause problems if you tie up the default dispatcher with blocking SQL queries while using clustering: the cluster gossip won't be able to fire and if your failure detection settings are low, that node can be marked unreachable by the rest of the cluster.