Hi Will,
I am working on a small framework called Thotti to define an to run such
tests. At the moment I isn't ready for public use, but I guess within
one month I could be used to run non-distributed tests.
Search for Thotti at Github. There is also an JIRA issue.
Bye,
Oliver
Am 07.11.2011 04:47, schrieb Weiquan Lin:
Hi,
Is there any performance comparison between different sizes of data
sets by using mahout+hadoop (with any algorithm), such as TB vs PB? or
10 million rows vs. 100, 1000, etc.?
Also, has anyone done any performance comparison between mahout vs.
non-mahout solutions?
Thanks,
Will
