Download the authoritative guide: Cloud Computing 2018: Using the Cloud to Transform Your Business
As the Head of Technical Strategy for Rackspace International, Toby Owen is responsible for ensuring the organization has the technical capabilities to cover current and future roadmap commitments, and leading the technical go-to-market strategy for new International markets.
In this interview, we spoke about the most common issues companies deal with as they migrate to the cloud, including choosing a vendor and understanding the importance of service agreements.
Toby Owen, Rackspace
1) When a company looks for a cloud computing vendor, what are some key questions they should ask?
One item to understand is whether your cloud vendor has a complete set of features. Everyone has virtualized servers (aka compute), but what about database services? Load balancers? Deployment automation capabilities? Do ALL the services have API access. In other words, will they work well in a fully automated deployment and management capability, where the application itself can participate in configuring the services to be used?
Another item to consider is portability. Does your prospective cloud hoster use proprietary technology, which will make it difficult to move your application elsewhere if things don’t work out? Or do they support a standard where that application would be easier to move if needed?
2) What applications are ideal for the cloud? Which apps are better served by being hosted in-house?
This is probably the most important question for companies just beginning to explore cloud services to answer. Understanding how applications can benefit from the cloud helps with understanding which apps are best suited to it. Apps with high fluctuations in traffic can take advantage of the utility billing of cloud, saving money by only deploying extra capacity when needed rather than leaving server capacity idle.
Apps built to leverage the built-in automation features (APIs, automatic scaling, etc) are also best fits for clouds because the application itself can trigger new capacity when needed, efficiently use cloud based storage and so on.
Apps that account for resiliency and high availability as part of the application are best, since the cloud is generally built from commodity hardware. It can scale out (horizontally) very fast but depends less on built-in hardware resiliency, relying on applications to account for scaling and fail over needs.
And of course, development and test environments also are a great fit, since the rapid deployment times for a new development environment can really speed up dev cycles. And the ready access to massive scale for load testing also makes full load testing economically feasible and relatively quick, as opposed to paying a vendor to load test for you, or waiting for the IT department to provision extra servers.
3) Should companies move all at once to the cloud? Or more gradually?
It’s always best practice to get familiar with a new technology rather than jump in headfirst. The same goes for cloud. Since there is a fundamentally different approach required for architecting applications to fully maximize the economic and rapidly elastic benefits of cloud computing, an understanding of your own application architecture will help tremendously when identifying which workloads will be best to migrate first.
Choose ones that allow relatively simple migrations, such as web apps, where the web and app server tiers are typically already built to scale horizontally. Move these components first and allow yourself time to get familiar with how to interact with the API and dashboards your cloud provider makes available. Tackle the complex data management apps once you are more experienced with how to scale out and you know the performance characteristics of cloud infrastructure.
4) What about the hybrid cloud? How should companies view this concept?
Hybrid cloud is a combination of public cloud resource and other infrastructure types – private cloud, bare metal hosted servers or even your own datacenter infrastructure. By combining multiple resource types, you are able to fully realize all the benefits of the cloud while still maintaining stringent performance and security requirements.
For example, Rackspace offers a solution called RackConnect, allowing you to pair dedicated bare metal servers with cloud servers, which can be optimal for pairing web and app tiers of your app with cloud’s rapid scale-out capabilities, while maintaining a single tenant database on high performance bare metal servers.
5) What about service and the cloud? Given that choosing a cloud vendor may be a multi-year commitment, how should a buyer think about the service component of a vendor’s offering?
Service is always a critical aspect of any vendor relationship. With cloud computing, and particularly with hybrid cloud, complexity can be daunting. Some customers are happy to do it all on their own, but most who turn to the cloud are not experts at cloud. They are experts in their business and optimizing their applications to improve and enhance their business, and taking time away from this mission to learn a new technology stack can be counter-productive.
The technology of cloud is evolving so rapidly, and the choice and variation so vast, that partnering with a provider that can be your trusted advisor is more important now than ever before. Interview all your potential partners and work to understand the depth and breadth of their offerings. What will be their response if your application suddenly goes offline? Can they help you assess which applications would be good candidates for the cloud and build a roadmap with you for migrating them to the cloud? And, can they help you continue to evolve your business at the pace of technology change? All these factors are critical and will ensure you have a successful journey to the cloud.
BIO: Toby Owens has been with Rackspace for 4 years, and has over 17 years of experience in IT across leadership, product development, technical and operational roles. Prior to joining Rackspace, he led the Web Services team for Wells Fargo Bank with responsibility for internet banking infrastructure and support. He holds a bachelors of science in engineering from the University of Oklahoma, and an MBA from the University of Texas in Austin.