A feature-complete release candidate of the 1.0 release is now ready. This is a foundation you can build on with confidence.
The topic: "Scaling by Cheating: approximation, sampling, and fault-friendliness for scalable Big Learning"
New "tag" API, faster Hadoop jobs, fixes and other improvements in this final beta version
Myrrix and NGDATA work together to integrate Myrrix as part of Lily
The Myrrix Computation Layer works seamlessly with the Hortonworks Data Platform
Now with clustering, new API methods and integration points. A 1.0 release candidate.
Top video sharing site adds 600,000 views per day with Myrrix
Many fixes and improvements from the field, in advance of the 1.0 release
Access the Serving Layer REST API from PHP
There's a nice mention of Myrrix here from Amazon's Spotathon contest ...