Thanks, this is really helpful.

On Fri, Mar 4, 2011 at 5:15 PM, Danny Bickson <[email protected]> wrote:
> I perfectly agree. And I just fixed the graph to plot the full range of
> values.
> Thanks for the confirmation about experiment setup. If anyone has any tips I
> would love to hear them further.
>
> - Danny
>
> On Fri, Mar 4, 2011 at 8:02 PM, Ted Dunning <[email protected]> wrote:
>
>> I generically dislike graphs with offset references.
>>
>> Your run-time graph has a baseline of 1500 seconds which makes it tricky
>> for the reader to understand your (entirely correct) statement that having
>> more than 8 machines isn't helpful.  The way you plotted the graph
>> coincidentally looks like nearly perfect speedup across the entire range.
>>
>> No comments about your setup.  My guess is that you could tune hadoop to
>> get a better result due to lower overheads but the results won't be
>> categorically different.  Iterative algorithms on stock hadoop are just
>> plain problematic.
>>
>>
>> On Fri, Mar 4, 2011 at 4:33 PM, Danny Bickson <[email protected]>wrote:
>>
>>> I would love to get any feedback you or others may have about the setup of
>>> this experiment.
>>>
>>
>>
>



-- 
Lance Norskog
[email protected]

Reply via email to