VoltDB 1.2 released earlier this month:
New data availability features. Version 1.2 introduces two important data availability enhancements. The first is network partition tolerance, which allows VoltDB to automatically detect, isolate and manage network failures. This is a critical feature for distributed database infrastructures including those deployed into public clouds such as Amazon’s EC2. The second availability feature, node rejoin, allows VoltDB database nodes that have been taken offline (e.g., for maintenance or repair) to “rejoin” the cluster while the database is live. Node rejoin dynamically resynchronizes all node data.
I’d love to read more about about the mechanisms used for automatically detecting, isolating and managing network failures. (If I remember correctly) The topic of reliably determining partitions in a distributed system is a central part of Seth Gilbert and Nancy Lynch paper on CAP theorem. It would also be interesting to understand how VoltDB deals with its strong consistency promise in these situations.
And some management tools (nb: by the announcement text I cannot tell if they are available only in the Enterprise version):
New consoles for provisioning, management and monitoring. New in the Enterprise Edition of version 1.2, the VoltDB Enterprise Manager (VEM) provides database and systems administrators with browser-based tools for managing production VoltDB databases. VEM offers a flexible suite of consoles for performing many common administrative and diagnostic activities.
Original title and link: VoltDB Release: Version 1.2 Featuring Data Availability Enhancements (NoSQL databases © myNoSQL)