Hoping for a more cost-effective and flexible IT infrastructure, many enterprises pursue a transition to the cloud. By examining different cloud models, such as software as a service and infrastructure as service, IT teams can understand common uses for cloud computing, as well as the necessary administration skills the cloud requires.
Whether you use the cloud for bursting, disaster recovery or a number of other capabilities, a cloud computing deployment affects everything from IT capacity planning to workload management. Here are some tips for data center teams transitioning workloads to the cloud, and considerations for future cloud use.
Weigh public cloud vs. colocation options for DR
Before choosing public cloud or colocation for disaster recovery (DR), recognize your organization's needs. Fast response time is crucial for either public cloud or colocation, but there are differences between the two options. With a non-managed colocation service, organizations will rent or lease the facility space, with the IT team controlling full DR policies. A managed colocation service provides the IT equipment and facility space, as well as monitoring, management and DR services.
When using public cloud for DR, the flexibility is often greater. Major cloud providers, such as Amazon Web Services, can support a large number of DR policies set by your IT team and integrate them into a larger plan. The transition to cloud for disaster recovery is often cheaper, as well, since organizations only pay for the resources they consume.
Smaller businesses find the cloud model more appealing because of this flexibility and an ability to make changes quickly without being locked into a long-term plan.
Know what SaaS means for IT
Before making the transition to the cloud with software as a service (SaaS), weigh the pros and cons of the move. As one of the most common cloud computing models, SaaS offers the opportunity to maintain, customize and control software on infrastructure owned and operated by the SaaS provider.
Major benefits of SaaS include ease of access and reduced capital and operational expenses. SaaS enables IT to transition from an operations to a facilitation role, and reduces the budget and staff necessary to run software. In addition, SaaS speeds up the process from purchasing a service to seeing the impact on your bottom line. The SaaS is responsible for redundancy, repair and infrastructure scaling.
Despite these benefits, some drawbacks of SaaS involve the remoteness of SaaS applications. Bottlenecks and other issues, such as performance variabilities, also affect the usefulness of SaaS.
Some of the best SaaS examples are suited for commodity applications, such as email, file sharing and expense reporting. New applications with high customization levels would not be recommended for SaaS, since cost benefits may be lost with implementation.
Explore the role of IaaS
Another popular cloud model, infrastructure as a service (IaaS), allows organizations to purchase virtual infrastructure, including servers, storage, network services and databases, that is hosted by a cloud provider. With IaaS, the transition to cloud replaces parts of an organization's internal data center infrastructure with a hosted component.
The best workloads for IaaS hosting are those running on outdated, on-premises systems that require an upgrade. Also, applications that have a wide range of user demands can be handled well on IaaS.
For organizations considering a transition to the cloud, some real-world IaaS examples include scaling externally facing applications, improving consumer-facing content, disaster recovery and application development and testing.
While IaaS can be cost-effective, it isn't always. For example, applications that are heavily dependent on legacy processes can be a major headache to overhaul and run in IaaS.
Consider a cloud bursting architecture
Hybrid clouds offer organizations the opportunity to develop a cloud bursting architecture. But after making a transition to the cloud, a lack of network bandwidth to handle these cloud bursts becomes a significant problem.
To avoid cloud bursting issues, businesses can choose to run IT fully in the public cloud, use a colocation provider with a faster wide area network, build an on-premises cloud bursting architecture or set up a wall where some data always remains in the public cloud and the rest lies in another form of cloud storage.
Security and cost considerations affect which possible cloud bursting architecture a business chooses to deploy. Outsourcing workloads with pertinent information can make IT uneasy, even with counter-threat measures and added security features. Concerns about costs and performance are also likely to occur, as data stored off premises, although often cheaper, can take longer to run and impact business decisions.
Evolve IT capacity planning for cloud
The future of the data center continues to shift, and the transition to the cloud will affect what lies ahead for IT capacity planning. The future will see a combination of traditional architecture with a cloud interface and will force organizations to handle the growing demand for increased speed and efficiency in data transfer. The emergence of hybrid cloud has altered how IT views and performs capacity planning and, as data center hardware continues to improve, organizations have to decide whether to stay in-house or send their information elsewhere.
An on-premises private cloud keeps everything with internal IT, but the learning curve and subsequent costs to maintain that cloud can be a major turnoff, especially for small and mid-level organizations.
Instead, these businesses will most likely turn to public cloud for the lower costs and ease of use -- which means fewer in-house servers to manage. Other technologies, such as containers and in-memory processing, will also reduce the number of servers an organization needs.
How are colocation and DRaaS different?
IaaS cloud consolidation continues to grow
Expand the data center outside DC walls