Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'd like to get some advice on what factors influence the decision of how large (in terms of resources) you would decide to size each node of a virtual machine cluster (VMware, Xen, etc), and in what way. What are the benefits/drawbacks of multiple smaller individual nodes (less cores, RAM, etc) vs. a smaller number of larger individual nodes? Is there a balance to be struck, and if so, in your experience, what do you consider when attempting to find the optimum?

I appreciate that there almost certainly isn't a one size fits all answer to the question, but what factors would you take into consideration?

share|improve this question

closed as not constructive by Shane Madden, warren, HopelessN00b, Michael Hampton, John Gardeniers Nov 23 '12 at 8:43

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.If this question can be reworded to fit the rules in the help center, please edit the question.

This is a one-size-fits-none question as it's entirely dependent on the kind of workload you intend to run on the cluster... – voretaq7 Sep 9 '11 at 14:56
Ok, I was happy to take an answer along the lines of "X & Y type workloads would influence the decision in Z manner." That's why I worded the question as 'what factors would influence.' Suppose it was a general purpose cluster. A good example would be a cloud hosting provider for mainstream use. – dbr Sep 9 '11 at 15:18
It'll come down to cost. If you can afford the costs associated with lots of 4-socket servers, then do it. But the better bargain is smaller 2-socket systems. – ewwhite Sep 9 '11 at 15:26
up vote 0 down vote accepted

It is a factor of workload. But these days, licensing dictates a lot of it. Most of the VMWare installations I've performed have been with smaller organizations, so I had to optimize the number of cores per CPU socket. These days, a good 2U server with plenty of RAM expansion (e.g. HP ProLiant DL380 or Dell R710) is a good bet. Those systems are optimized for virtualization. It's the sweet-spot in terms of price/performance. You'd need to provide more details on what you are looking to do in order for us to give a better answer, though.

share|improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.