Here is a nice summary of the Tango paper.
<a href="http://muratbuffalo.blogspot.com/2014/09/paper-summary-tango-distributed-data.html" rel="nofollow">http://muratbuffalo.blogspot.com/2014/09/paper-summary-tango...</a>
"the abstraction of a replicated, in-memory data structure (such as a map or a tree) backed by a shared log"<p>If I read just this piece of text anywhere, the word popping up in my mind would be zookeeper
A couple of my friends have been looking at this paper and created their own visualization implementation:
<a href="https://github.com/derekelkins/tangohs" rel="nofollow">https://github.com/derekelkins/tangohs</a>
Why need a shared log? Remember the CAP theorem. No need for these bottlenecks. If you want to store that A happened after B, just have A store a (hash of) B.