Written by Andrew Wilson.
VoltDB is evolving constantly and we have just released a our own version of the Nodejs-VoltDB driver based upon Jacob’s excellent work.
Some of the new features include:
- Cluster support – the client can connect to all VoltDB servers and round-robin queries across the cluster, yielding faster performance.
- Connection failover – if a VoltDB server goes down then the client invalidates the connection and routes new queries to working connections.
- New client interface providing an easier way to connect to and query from VoltDB.
- Query conversation management that will allow read operations when they would normally be blocked by looping query requests.
- Supported by VoltDB.
We’re writing code and unit tests, with the goal of elevating the Node driver to be on par with our Java and C++ drivers. We’re also building some throughput and scaling tests, so stay tuned for more performance info as that data becomes available.
What’s the Fit Between Node and VoltDB?
As I mentioned above, both Node and VoltDB are optimized to support asynchronous, event-driven apps. VoltDB can deliver pretty amazing throughput and scale, particularly when driven asynchronously (read a bit more about Volt’s throughput and scale here). It also supports ACID transactions. So Node developers should consider using VoltDB for any performance-sensitive app, particularly when data consistency is also a priority.