KML_FLASHEMBED_PROCESS_SCRIPT_CALLS
 

How To Optimize Cloud Server Workloads to Maximize Efficiency

September 24th, 2012 by - 6,001 views

If you’re familiar with cloud infrastructure and infrastructure-as-a-service (IaaS), you probably understand the substantial benefits that come along with deploying infrastructure in the public cloud: things like “utility billing and on-demand availability,” “elastic benefits that let you scale resources up and down based on demand,” and “the ability to rapidly move and redeploy workloads as needed.” This flexibility is why we originally brought GoGrid’s hourly pay-as-you-go Cloud Servers to market. They’re perfect for specific cases like these:

  • Periodic workloads that only run for a few hours, days, or weeks during a given billing cycle
  • Short-term, project-based workloads where term commitments aren’t desirable
  • Short-term spikes in workload where demand is erratic and being able to scale resources up and down quickly are desirable
  • Development and test workloads that require rapid iteration and redeployment of resources
  • Proof of concept workloads where instant access to resources and the ability to quickly change technology are key

Customers with steady-state and long-term workloads don’t always need this hourly flexibility, however. And that’s why GoGrid has developed prepaid monthly, semiannual, and annual Cloud Server products. Prepaid Cloud Servers are less flexible, but they do offer significant cost savings in exchange for the term commitment. The shortest prepaid term GoGrid offers is a monthly prepaid Cloud Server and the longest term is an annual prepaid Cloud Server.

If you run a constant workload during a given month, a prepaid term server is probably a better solution than an hourly server. Again, the tradeoff here is flexibility. Prepaid servers are ideal for:

  • Steady-state workloads where demand is constant
  • Workloads that tend to grow rather than contract
  • Production applications where you can plan for demand in advance

For example, imagine you run an eCommerce website. You know you always need three servers to run your operations throughout the year. During the holiday season, however, you know demand is likely to spike. Your deployment of annual servers going into the holiday would look something like this:

Chart1

As the holiday season starts, you see an increase in demand that is unlikely to drop off, so you ramp up the number of monthly Cloud Servers deployed. You plan to meet demand as it increases, and your usage now looks something like this:

Chart2

With the money you’ve saved by deploying prepaid servers, you’ve been able to plan two big holiday promotions. One promo kicks off the last week of October and the other the second week of December. These promotions are more successful than expected, creating massive traffic on your website. Day by day you monitor demand, deploying hourly servers as demand increases and scaling back as each promotion winds down. During this critical time for your business, your usage looks something like this:

Chart3

The success of your holiday season has created such a positive customer experience that it has increased your long-term workload. You now have the confidence to deploy additional annual Cloud Servers or keep a few of those extra monthly Cloud Servers online. Because you’ve learned how to tailor your workload demands to the type of server you use, you’ve managed to save money and handle your budget more effectively.

To determine how you can benefit from our new Cloud Server offerings, simply take a look at your workloads and ask yourself if there’s room for optimization by server type. For specific details on our server pricing, please visit the Cloud Server product page.

The following two tabs change content below.

Kole Hicks

Senior Director of Product Management at GoGrid
Kole Hicks is the Senior Director of Product Management for GoGrid, the leader in Open Data Services (ODS) and committed to delivering purpose-built, non-opinionated Big Data solutions and services for the management and integration of open source, commercial, and proprietary technologies across multiple platforms..

2 Responses to “How To Optimize Cloud Server Workloads to Maximize Efficiency”

  1. Very nicely you explain in this blog. I like the way that you choose to tell us. The graphs are very nicely used by you in this blog.

  2. I really like the example you used in this blog post regarding increasing the usage of cloud servers during periods of high website traffic, like the holiday season. The graphs are also a very nice touch in illustrating your point.

Leave a reply