@juanjovazquez thanks for the kind words :smile:. We have already implemented all tickets except one ( #142 which is rather easy) needed for an Akka cluster integration ( #143 ). I think that usage of ConstructR wouldn't even affect the ticket but I need to take a closer look. In general, using coordination services that are on the CP side of CAP (like etcd and consul) can only be relevant for single Eventuate location (= availability zone) but not for coordinating a distributed, multi-location Eventuate application which is on the AP side of CAP. Furthermore, an Eventuate - Akka Cluster integration should deal with the fact that group communication within an Akka cluster is not on the CP side of CAP (i.e. it doesn't use a consensus protocol). This can lead to split-brain scenarios where you can end up with two leaders, for example. The Eventuate - Akka Cluster integration will be able to deal with this situation without giving up availability and rather resolve write conflicts when the partition heals again.