Reasons frequently cited for adopting a cloud strategy include reduced time for application and infrastructure provisioning, flexible capacity, lack of internal technical resources, the benefit of variable costs, and a reduction in total cost of ownership, among other business-specific drivers. But if your reasons for adoption include cost and efficiency parameters then you should also be cognizant of potential costs amplified by a multi-solution approach.
[ALSO: Top 10 cloud tools]
Multi-vendor cloud deployments impose additional complexities and administrative overhead, including the need to:
- Manage multiple contracts and business relationships.
- Monitor and enforce multiple service-level agreements.
- Monitor systems performance, conducting root-cause analysis on issues and applying resolutions.
- Manage multiple charge-back schemas if you’re passing costs on in your organization.
- Perform audits on multiple cloud service providers to meet compliance requirements.
There are also technical complexities to consider:
- Managing resources across multiple providers often requires the use of third-party services that provide a consolidated management platform for server and application monitoring, access controls and policy enforcement.
- Programming with multiple APIs or employing costly API middleware to create integrations among applications.
- Network connections to multiple cloud providers must be managed to ensure peak performance and security.
- And using multiple providers means maintaining multiple network management schemas.
Security is often a primary concern when adopting cloud services and the multi-vendor approach poses some additional challenges:
- Managing multiple security schemas introduces complexity and risk and can require additional resources to monitor and manage.
- Single sign-on capabilities among disparate cloud providers may require additional third-party services or create additional complexities on developers.
- Security managers may incur additional overhead for enforcing security policy and meeting compliance requirements.
One of the key benefits of standardization on as few cloud services as possible is limiting the variety of environments you manage, thus reducing your administrative overhead. Minimizing this overhead leads directly to operational expense savings. And when you are dealing with a single (or a few) provider(s), your staff becomes well versed in the environment, which enhances IT agility, speed of service delivery, elasticity of services and acceptance of new systems.
Cloud services now dominate IT deployments for most CIOs who are cost, value and time conscious. As the CIO of a large and complex research university I’m responsible for the technologies that operate many businesses inside the institution, such as transportation, food service, hotels (residence halls), healthcare, public safety, sports entertainment, and of course, education and research. As is certainly true for other business organizations, the ideal of using a single cloud service provider is probably unrealistic, but it is still ideal to strive to use as few suppliers as possible.
At Georgia Southern University we’ve standardized largely on Google as our primary cloud service provider. Our strategy is to leverage Google’s capabilities to provide cloud-based applications, compute and storage capabilities that serve the diverse needs of the university. We do leverage other cloud services, but only after careful consideration of the many issues, complexities and requisite resources associated with adding providers.
Our experience hasn’t been without difficulties or tense moments, but as a result we’ve significantly reduced costs, improved service availability, and fostered an environment of innovation that leads to improved services for our students.
While concerns about vendor lock-in in a single-provider scenario are legitimate, these concerns can be effectively managed and planned for:
- Choose vendor-agnostic applications that will run in many cloud vendor infrastructures, which will help keep your options open. Open source solutions often provide excellent levels of functionality and reliability. Moreover, they’re portable among many cloud providers.
- Have an exit strategy and make sure that strategy and any associated costs are part of your continuity of operations plans.
- Consider where your data resides and if the provider utilizes data replication services across multiple sites and infrastructures.
- If you deploy instances in multiple data centers of a single cloud provider you can realize comparable redundancy you might obtain from multiple cloud providers.
- Negotiate a performance-based contract with solid service-level agreements you can live with. Take an honest look at your business. Can you live with a minor outage? You probably did before you consolidated services in the cloud.
- Write contracts that provide clear dispute resolution agreements. Taking the time to build a clear and favorable language helps mitigate risks associated with billing disputes or a disagreement over the interpretation of a contract that can cause rifts in your relationship with a cloud provider, which may lead to service disruptions.
- Work with your network and cloud provider to optimize the route between your points of business and your service provider. Discuss and document up-front what the capabilities are for optimization and alternate path utilization.
Even if your organization doesn’t adopt a single-vendor approach, but rather limits cloud service partners, the concepts presented here will help ensure you’re making contributions to your organization's bottom line while limiting costs and mitigating risks. For Georgia Southern University, this strategy has helped enhance services, reduce costs and improve graduation rates for students. That’s something everyone can be proud of.
Steven C Burrell, Vice President for Information Technology and Chief Information Officer, Georgia Southern University