data vault 2.0 has been in the lab for over 3 years now. it is time to evolve, to offer some pain relief to those of you struggling with joins, performance and partitioning on the physical level. dv2.0 is a specification that solves a few of the issues that are brought on by the strict […]
second look at technology stacks in the hadoop, mapreduce arena, and beginning to consider the technology to implement for data vault model testing.
i’ve long maintained that the data vault architecture is backed by mathematics, well, here’s a post that explains briefly the architectures and principles at work, including mpp, vertical and horizontal partitioning, and shared nothing components. we discuss why the number of joins in the model are a moot point.