Jobs such as generating and updating the crawlDB are bound by CPU and in local mode it's not taking advantage of more cores.
On Tuesday 28 September 2010 14:15:10 ramires wrote: > I think nutch with hadoop is very slow. Standalone nutch (1.2-rc4) with > HugePages and just html parser (tika parser very slow) it becomes a > rocket... > Markus Jelsma - Technisch Architect - Buyways BV http://www.linkedin.com/in/markus17 050-8536620 / 06-50258350
