Reserved Capacity vs Usage based Capacity

As a cloud broker, Gravitant works with clients to match their needs with what different cloud providers have to offer.

In this process, we have noticed that most of our clients have high transaction volumes on their Dev and Test environments.  However, in the Production environment, transaction volume is low with a lot of variability.  This makes sense because transaction volume in the Production environment is customer driven.

And most of the cloud providers we work with have lower rates per hour for dedicated capacity and higher rates per hour for usage based capacity.

Therefore, the lowest cost solution for most of our clients is to go with dedicated capacity for their Dev & Test environments, and usage based capacity for their Prod environment.

However, this may not be the case for all clients.  Therefore, it is important to analyze historical transaction volume and utilization for each environment and application type separately in order to identify the optimal combination.  Eventually, the main objective is to derive the optimal target of reserved capacity for each environment/application.  More on this topic coming soon...

Acknowledgements:
I would like to thank Robert Jenkins, CTO of CloudSigma, for his input to this discussion.

Free app for Cloud Consulting!

Gravitant has just launched their Cloud Value Planner free app for Cloud Consulting...

Customers can now simulate the capacity and cost impact of replacing their current servers with a private or public cloud.  Consultants can also use this tool to show potential customers the value of going cloud.

Check it out here...

Comparison of Traditional Hosting and Cloud Computing Solutions

Cloud Consulting in an app?

Researchers at Gravitant have successfully converted their Cloud Consulting practices into an app that will soon be available on the company website.  The app, which will be called the Cloud Value Planner, allows users to simulate the capacity and cost impact of replacing their current physical servers with virtual private cloud or public cloud options.





“The Cloud Value Planner is like consulting-in-a-box.  Questions that typically take 60-90 days and thousands of dollars to answer are solved within a few minutes using this app.” – Mohammed Farooq, CEO Gravitant

This simulator, which is still in its beta phase, will be launched as a free app before the end of February 2011.  Please check www.gravitant.com for updates.

Stay tuned for details on specifics of the app…

Jobs phased out by the Cloud

Is your job secure from being phased out by cloud computing?  Cloud providers will be responsible for most of the technical requirements, so where does that leave you?

Most technical issues in computing will soon be handled by the hypervisors, load balancers, etc. on the suppliers end. However, each supplier will have to be monitored and managed through service contracts.

Infoworld has specified some role shifts based on Forrester and Gartner:

IT jobs of the future will be highly administrative with a focus on capacity planning and contract management – Gravitant, Inc.

Quick tools to help you prepare for this shift are available here…

Overutilization vs Underutilization of Virtual Machines

Does virtualization truly improve utilization?

If so, shouldn’t cost be lower as a result?  Well, maybe in the short term…  but in the long term we are starting to see cases where costs accumulate over time and virtualization ends up being more expensive.

Gravitant’s response:
“Over-provisioning of Virtual Machines has resulted in Virtual Machine sprawl, which is difficult to manage.  This has resulted in higher cost over time.  On the other hand under-provisioning results in lower performance and SLA penalties.” – Mohammed Farooq, CEO Gravitant

Gravitant’s capacity planning tool configures the optimal physical and virtual machine landscape of servers, network, and storage to

1. Minimize VM Sprawl
2. Meet application performance thresholds
3. Reduce data center costs – systems and operations

Operations Research tools are used to balance the tradeoff between cost (due to sprawl) and performance (from SLAs).

3 keys to Capacity Planning for Virtualization

Fact:  Everyone’s going virtual to reduce cost and improve utilization. Why not?  Sharing of resources and paying for resources on-demand should be beneficial.  But that just makes the job of the capacity planner even more difficult!

3 big challenges in Virtualization that do not exist in traditional capacity planning:

1. Capacity of each box is dynamically allocated, so which virtual machine (VM) actually got how much of resources?

2. Each box has overhead utilization from each VM due to the hypervisor which reduces performance, so what is the critical number of VMs to be configured on each box?

3. Cost models are complex with options for on-demand vs dedicated vs burst capacity, so which option should be chosen?

Solutions: 1. We use a couple of key performance metrics (Transaction Rate and Response Time) that are uniform across all layers and applications.  This tells us how much capacity was effectively used by each application.

2. We use a slowdown factor that discounts available resources due to hypervisor utilization.  As a result, we can derive the optimal number of VMs on each box.

3. Because of solutions 1 & 2, we are able to accurately forecast capacity requirements which can then be compared with the different cost models.  If capacity requirements are high but stable, dedicated would be cheaper, but on-demand is better for unstable capacity requirements.