Towards Fast Invariant Preservation in Geo-replicated Systems

Today’s global services and applications are expected to be highly available, scale to unprecedented number of clients and offer reliable, low-latency operations. This can be achieved through geo-replication, especially when data consistency is relaxed. There are however applications whose data must obey global invariants at all times. Strong consistency proto-cols easily address this issue but require global coordination among replicas and inevitably degrade application through-put and latency. While coordination is an inherent requirement for maintain-ing global application invariants, there are instances where coordination on a per operation basis can be avoided. In particular, it has been shown that either moving coordina-tion outside the critical path for executing operations, or having one coordination round for multiple operations, are both effective ways to maintain global invariants and avoid most of the penalties of coordination. Our stance is that expanding this idea to geo-replicated settings has yet to be fully realised. In this paper, we review the design space of current solu-tions for engineering geo-replicated applications and present our guiding vision towards a general technique for providing global application invariants under eventual consistency, as a much cheaper alternative to strong consistency. 1