 |
Ted Dunning |
at Oct 20, 2007 at 4:39 pm
|
⇧ | 
| |
That is a pretty small amount of memory.
It should work, but you will see odd performance problems with some programs
due to memory binding.
Hadoop is generally very good about using hardly any memory at all, but it
is common to write programs that collect a fair bit of data during the
reduce step. That could easily be a downfall for you.
On 10/20/07 7:04 AM, "Sukalyan Banga" wrote:
Hi!
I am new at clustering, to start a demo cluster what are the steps require
to take.
Presently I could experimenting with 3 Linux system all latest Intel based
with 512 RAM
Thanks
Sukalyan Banga
