Cascading 2.6

We are happy to announce that Cascading 2.6 is now publicly available for download.

This release contains new features and bug fixes.

Of note are the new DecoratorTap and DistCacheTap (itself a DecoratorTap sub-class) classes. Working together, Flows can cache data directly into the Hadoop distributed cache when accumulating data for a HashJoin.

And for Driven users, new Java annotations allow for additional meta-data to be sent to the Driven UI when visualizing assemblies. For example, the ‘expression’ in the ExpressionFilter will be visible in the Driven 1.1 EAP.

What’s New in Cascading 2.6:

  • Added c.t.DecoratorTap class to simplify wrapping a given c.t.Tap instance with additional meta-data.
  • Added c.t.h.DistCacheTap a decorator for a c.t.h.Hfs instance that uses o.a.h.f.DistributedCache to read files transparently from local disk. This is useful for c.p.HashJoins.
  • Added new Java Annotations for managing the visibility of data sent to a remote management service.
  • Updated Apache Hadoop to 2.4.1 in cascading-hadoop2-mr1
  • Updated c.p.a.AggregateBy and c.p.a.Unique to count cache flushes, hits, and misses. Previously only AggregateBy tracked cache flushes.
  • Fixed issue where c.p.a.AggregateBy and c.p.a.UniqueBy were not honoring the c.t.Hasher interface

For more details on new features and resolved issues see: