On Dec 11, 2007, at 12:27 PM, Rui Shi wrote:
We have a lot of legacy code written in C++, Perl... I was wondering what is the best practice to leverage those legacy code in hadoop? It seems that HadoopStreaming could be a choice. But I can not find more information about that.
Streaming is easy for the scripting languages and for debugging, especially if you data is textual. You can get a richer interface, if you use the C++ interface: http://tinyurl.com/2zvyfa
-- Owen
