I was just debating the basic model of this recently (not knowing of Amazon's work) and how it could be made to operate if we could relax consistency requirements. Makes me happy to hear that smarter minds than mine are way, way ahead of me and building software around this.
Anyone know of accessible implementation of this for one's own clusters or EC2 machines? I assume SimpleDB isn't built on Dynamo?
Hmmm... I suppose S3 is pretty much is built on Dynamo? Let me read up on it, but if other people know for sure, I'd love some of your insight.
the paper is heavy on the insights but less so on the implementation, as to be expected. very interesting and innovative system for its problem space.
also, there are many moving parts to this system. if you read through the paper odds are that you would find parts applicable to your problem. pluck some of the parts for use in your own system.
Anyone know of accessible implementation of this for one's own clusters or EC2 machines? I assume SimpleDB isn't built on Dynamo?
Hmmm... I suppose S3 is pretty much is built on Dynamo? Let me read up on it, but if other people know for sure, I'd love some of your insight.