by Karthik Ramarao, CTO APAC, NetApp
What does a typical mall food court have in common with IT infrastructure options of today? More than you might think upon first glance; the obvious choice may not always be what it seems. This analogy may seem far-fetched, but it can actually serve as a guiding light for thinking through IT strategy approaches and how they can help organizations face today’s data challenges. As with any form of sustenance, there are a vast number of choices available for implementing, securing and managing data both internally and externally.
As Jay Kidd pointed out in his 2014 predictions, 2014 will be the year when a large portion of businesses transform their IT operations from a “build-operate” model to a “broker-provider” of IT services model. In addition, there is probably no hotter topic in IT conversations right now than cloud. Organizations around the world are adding cloud to the IT menu with hopes of delivering IT faster, cheaper, and smarter, ultimately making their businesses more responsive and competitive.
However, as cloud environments become more popular, businesses are increasingly struggling to manage their data freely across dedicated and hybrid environments. Everyone wants to confidently manage and maintain control of their data across any environment, accelerating innovation and IT responsiveness, but the challenge is that this is much easier said than done. It requires dynamic approaches to infrastructure design. Ultimately, before a solution can be considered, it’s critical to understand the business needs and restrictions that will impact the choice.
The Food Court Analogy of IT: Do you feel like eating in or out?
To continue with the food analogy, the menu of choices for data infrastructure models is organized between dedicated and shared models. Choosing between the two is eerily similar to choosing between cooking or eating at home or going to a food court. Both offer a common set of amenities with a basic goal – managing your appetite / data.
The dedicated infrastructure model focuses on personalization, security and control, while the shared model’s aim is efficiency and cost. For organizations today, it comes down to the question: do the model’s benefits outweigh its risks/downsides? The same factors when choosing an infrastructure model need to be considered when choosing to eat in or out – cost, control, efficiency and governance.
Let’s boil this down a little further.
Who’s coming to dinner?
Everyone has a kitchen where you live. It probably costs anywhere from ten thousand to thirty thousand dollars or more to build before you could make your first dish. It doesn’t do anything unless someone with the time, knowledge, skills and resources uses it to produce a custom product.
Like the dedicated infrastructure model, the capital investment and operational expenses to build and maintain a kitchen are quite high. Additionally, its boundaries are finite and kept on premises – meant for use by a family or invited guests. Unless you’re willing to make an additional investment to expand; the capacity, capabilities and size is often set in stone. Even if you do, the changes can’t be made over night. However, it is often designed to meet your exact needs and allows you to maintain complete control over the items, where everything is stored, who has access and what’s produced. For example, if you or one of your guests has a food allergy, you can customize dishes to meet their needs.
The same can be said for dedicated IT infrastructure. The capital investment and operational expenses to build and maintain the infrastructure are significantly higher, and requires someone on premises with the skills to manage and maintain it. It is also more limited in the scalability and capabilities; usually meant for one application or use by one department – although that is changing a bit. That aside, this model allows for complete control over your data, who has access to it and which workloads are deployed in it. Additionally, like creating a dish for someone who has a special dietary need a dedicated infrastructure allows you to manage data according to specific governance or compliance regulations the company must adhere to.
While both the dedicated infrastructure model and the home kitchen afford you complete control, they come at a price – namely time and money.
Would you like files with that?
The food court on the other hand, is all about flexibility, variety and efficiency. People don’t eat at a food court because they want a 5-Star dining experience usually. They eat there because it’s economical. Whether in Toronto, San Francisco or Singapore, food courts around the world share a set of common attributes. They are quick and easy, inexpensive, offer a variety of options and can serve one person or one hundred. Most importantly, you pay for what you want. But there are also drawbacks – namely the lack of control over the food, and difficulty of meeting individuals’ dietary restrictions.
It is the same attributes that draw companies towards shared infrastructure – private and public cloud environments alike. A company building a shared IT model for its various departments may spend less to build and operate what each of the departments may have spent if they individually stood up an IT for their respective needs. Services can be deployed quickly and scaled up or down as needed. Like food courts, public clouds services operate on a pay as you use services oriented architecture. However, lack of total control and customization means that only certain workloads and types of data can be moved to the cloud, in accordance with governance and compliance regulations.
IT starts with a balanced diet
When choosing either option – dedicated / home kitchen or shared / food court – it basically it comes down to control versus cost and efficiency. Thankfully, advancements in cloud solutions mean that IT departments are now seeing the benefits of dedicated infrastructure – greater control over their data and workloads – with shared infrastructure options.
Granted, there will be a need for dedicated infrastructure for highly sensitive or special workloads, but as cloud solutions advance and the total cost of ownership continues to drop, businesses will look to migrate many applications and workloads to the cloud.
Remember, as with dining options, the key to data management – and subsequently, business growth – is balancing needs with resources. Organizations don’t move to shared or hybrid infrastructures, data does, and data does not like to move easily.