This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
1. - Cloud computing decisions: Read more in this section
- Should data center managers consider cloud computing?
- Considering data center resource options
- Building up hybrid cloud with virtual data center services
- How to integrate hybrid cloud into corporate data centers
- How private and public clouds get along
- What deployment model works for your IT shop?
- Easing PaaS and SaaS integration with hybrid cloud computing
- Virtualization, cloud and SDN modernize data centers
Explore other sections in this guide:
- 2. - Data center costs in a hybrid cloud model
- 3. - Cloud management choices for the data center
- 4. - Hybrid cloud management tools
Careful planning and well-conceived usage policies must accompany hybrid cloud integration in the corporate data center.
Cloud computing couples the flexibility of virtualization with self-service, dynamic scalability and reporting that allow users -- rather than IT administrators -- to make compute decisions. Hybrid cloud computing couples a private cloud within the existing enterprise data center with workloads provisioned to a public cloud provider. Hybrid cloud computing for business workloads is evolving as cloud options mature.
Workload provisioning decisions
When planning hybrid cloud deployments, businesses decide which workloads stay in the private cloud and which go to the public cloud. This is one of the quintessential questions behind cloud computing in business: Where are workloads safe?
Any IT decisions should ultimately benefit and protect your business. You control the private cloud; you do not control the public cloud provider or WAN connectivity to it. This raises the specter of availability issues. Public cloud providers typically do not share workload performance or system efficiency details. The most mission-critical applications, such as transactional order entry systems or databases, usually stay in the private cloud under in-house IT monitoring.
Depending on the business, regulatory compliance can affect workload distribution. For example, a company subject to geographical computing or data storage limitations will keep business workloads in the private cloud rather than placing them in the public cloud where they might operate anywhere in the world. A growing number of public cloud providers can make certain regulatory concessions, such as only running workloads on systems within a specific geographic area. But the provider must still adequately certify its actions if the business wants to meet regulatory requirements. Consult with enterprise regulatory compliance officers and legal advisors on any public cloud workloads.
Consider data security and the business's obligations to regulators, shareholders, users, and other factions. Although no multitenant public cloud has had documented security breaches resulting in data loss or compromise, their service-level agreements (SLAs) always ensure that the provider is not liable for such loss, which might limit the benefits of cloud computing for business workloads. Sensitive data can be encrypted when in the public cloud with a high degree of safety, or even kept encrypted in the private cloud.
Workload performance is another provisioning issue: Where does the application or data work best? WAN bandwidth is typically limited, so workloads that require significant bandwidth -- e.g., many endpoints streaming content -- should stay on-site instead of on a public cloud.
Financial considerations can also influence workload provisioning. For example, a company seeking to trim its IT capital budget (CAPEX) and transition some capital costs to operating expenses (OPEX) may choose to move more workloads to the public cloud. This lowers the CAPEX on servers and other computing systems, allows a company to extend the working life of existing hardware platforms and reduces software licensing for system monitoring and other tasks.
Application programming interface (API) compatibility dictates how the workload interacts with the cloud environment and functions in the public cloud. Different public cloud providers have different APIs, and the IT team may need to recode workloads to get the full advantage of cloud hosting. This takes time and investment, and complicates workload mobility across public cloud providers. As a result, public-cloud-worthy workloads may stay on-premises because the business cannot invest in coding for various APIs.
Finally, consider data protection options. Private clouds offer various tools to protect workloads; some tools may already be available in the enterprise. Public cloud adopters should examine the provider's backup and restoration options. Workloads that require more aggressive protection and restoration may need to stay in the private cloud.
Managing the hybrid cloud
Workload provisioning speed and flexibility are some of the primary benefits of cloud computing for business operations. End users can provision and scale new workloads in minutes. Unfortunately, this kind of user control can also seriously harm any business unprepared to govern cloud usage. Many enterprises implement policies to regulate private cloud and public cloud use.
There is no single, ubiquitous hybrid cloud policy, but a coherent set of ground rules will save money and safeguard corporate interests.
Start with guidelines for initial placement of cloud workloads: Regulatory compliance, API compatibility, network performance and cost-effectiveness are part of the private vs. public cloud decision. Also ask: Is this workload needed?
Implement policies for reporting and usage reviews to ensure that each workload is monitored and its value to the business assessed regularly. With cloud self-service, users can establish a workload, use it for a while, and then let it fall into disuse. Even abandoned, the workload is still running, consuming computing resources, getting backed up and costing the company money. Sprawl is a very real disadvantage of cloud computing.
Application and usage awareness stave off virtual machine sprawl and other wasted resources. Unit managers must know who is running workloads, who is using them, when to migrate workloads between public and private clouds, when to scale computing resources up or down, and when to retire and reclaim the resources and budget for other applications.
Changing workload distribution in the cloud
As cloud use increases, companies are more willing to place sensitive workloads and data in public clouds. There is no doubt that databases, customer relationship management and enterprise resource planning applications, and other complex, enterprise-class workloads, are gradually shifting to the public cloud. The exact composition and purpose of each workload will vary, depending on the business size, regulatory obligations and industry segment.
TechTarget's 2013 Cloud Pulse survey data indicates that almost 62% of cloud users currently use some Software as a Service in the public cloud, such as Salesforce.com. Almost 56% of current public cloud users employ some form of Storage as a Service, and over 42% use collaboration software in the public cloud.
Public cloud for business use requires confidence. Businesses experimenting with public cloud computing quickly realize that their data and processes remain accessible, secure and managed. IT administrators, business executives and other decision makers are becoming comfortable and are committing more data and applications to the public cloud.
Public cloud services are not foolproof either. Service outages at the public cloud provider or disruptions to WAN connectivity can render public cloud workloads inaccessible. SLAs protect the interests of the provider, not the business. Public cloud customers have little real recourse after an outage, so consider worst-case scenarios when provisioning workloads.
Even as public cloud use grows, due to its flexible workload deployment and OPEX-focused resource model, private cloud in corporate data centers still provides fast, scalable and user-driven computing resources for the business. Integrating both cloud options into a hybrid cloud offers data centers the ultimate in workload versatility. Although most businesses keep mission-critical applications and data in-house, more of the company's hybrid cloud infrastructure will likely wind up in the public cloud over time.