Scale-up versus Scale-out

I just ran across a paper from IBM comparing scaling-up (using bigger boxes) to scaling-out (using more boxes). They use Nutch search as their workload, and conclude “… that scale-out solutions have an indisputable performance and price/performance advantage over scale-up for search workloads.” Not exactly a big surprise, but it’s good to have objective data. They also conclude that “Scale-out systems are still in a significant disadvantage with respect to scale-up when it comes to systems management.” Hmm. With frameworks like Hadoop, folks shouldn’t be bothered as much by the more frequent host failures that a scale-out system is prone to.


3 Responses to “Scale-up versus Scale-out”

  1. Mark Says:

    Interesting paper. Have you noticed the nose-dive on figure 11 for the Scale-out option at around 1000GB? It seems to be heading south of the scale-up figures. I notice they omit this measurement for the scale-out solution in figure 15 too.
    They don’t seem to offer an explanation for this.

  2. Mark Says:

    Withdraw last comment. I emailed one of the report authors and he pointed out the basis of my misunderstanding.The drop-off in performance was for the “scale out *in a box*” scenario where multiple Nutch instances are run on a single machine and memory usage was the issue.

  3. beautiful girl Says:

    beautiful girl

    Scale-up versus Scale-out | Free Search

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: