Hey Juan - There's actually an ongoing discussion about this. From my perspective it would make a lot of sense to extract the metastore/serde parts of hive as reusable components for use by all processing frameworks. Of course the Hive folks would need to agree with te vision of a common metadata and read/write path for Hadoop in general to do this.
If you have hive setup, even before this discussion plays out, you can read/write your hive data with pig today. The giraph folks also have an hcatalog adapter. I don't have a crystal ball to know how the hive/hcatalog stuff plays out, but from a user perspective it shouldn't matter much. Just use the hcatalog adapters and we'll take care of the rest! --Travis On Tuesday, October 30, 2012, Juan P. wrote: > Hi guys, > I've been reading the documentation and I was wondering if there's a plan > to merge HCatalog and Hive into a one project at some point. > > There seem to be some common goals, and a lot of dependency with Hive. > > I get the idea that Hive should strive to provide a lot of the > functionalities in HCatalog, and the latter needs a lot of the former's > infrastructure in order to work. > > Or did I completely miss the point? > > Thanks! > Juan >
