On 8/7/07, novice user <[EMAIL PROTECTED]> wrote: > > Hi, > In reduce phase, with outputValueGroupingComparator, we can sort all keys > and then group values of a particular key together and send it to reduce() > method. Is there a way to sort values of a particular key efficiently before > it reaches to reduce method?
I'm not sure if this is what you want, but Google's MapReduce framework has the concept of an optional second key parameter for subsorting of records (saw it in some slides by Jeff Dean). Perhaps you could integrate this into Hadoop as a patch and submit it? -- Toby DiPasquale
