Replication Publisher Infrastucture

[tsutomi]

Implement an infrastructure that subscribes to Data Definition events (eg. CREATE, ALTER, DROP, etc.) and publishes the commands to subscribers in a network (slaves).

[meikeric]

Have you considered a distributed concurrency framework? Something akin to Akka? There is a C# version, Akka.net. http://www.getakka.net. This would allow for the possibility of massively scaling out to multiple hosts.

[tsutomi]

@meikeric actually I didn't consider using an external tool for such: in fact, I was more focusing on developing an HDFS implementation of IFileSystem interface, to scale trough a cluster, but your ideas seems very valid: we could discuss it!