Introduction. For this week’s blog post we will be examining the concepts, benefits, challenges, and applications of a hybrid cloud data storage system. The primary sources for the referenced information are NetApp, Microsoft Azure, and Google – referenced below.
Personal Lessons. Reviewing the pros and cons of hybrid cloud systems helped me to conceptualize data storage and distribution through visualizing the various options. My main take-away is that each enterprise will have a different combination of internal cloud, external cloud, and on-site solutions which will maximize efficiency for that particular organization. The factors contributing to that unique combination include the enterprise’s business model (and the products of the enterprise), its size, and its security requirements.
Definitions. A hybrid cloud solution is a data storage solution which utilizes two or more of the following storage solution options: on-site, cloud private, cloud public. On-site solutions are owned and managed by the enterprise, close or on the premise of where they are applied. Private clouds are also owned internally by the enterprise, but are managed off-site – perhaps at a central facility serving muliple geographic locations. Public clouds are outsourced to providers, such as Google, Microsoft, Oracle, and a multitude of smaller businesses which offer B2B daata storage solutions.
Benefits of Each Approach.
A) Public Cloud. By utilizing a second-party, remotely, to manage an enterprise’s data storage requirements, a business can increase its flexibility. This is due to the inherent “renting” of space – when a company has higher requirements, it can subcontract more space quickly and easily from a supplier who specializes in storage space. When the business is no longer growing or shrinking, those subcontracts can be terminated. This coincides with a lower capital expenditure requirement for the enterprise – by outsourcing, physical infrastructure requirements are reduced. Another possible advantage is reliability – if an enterprise manages all its data in-house, and there is a problem with one of the components (personnel, physical resources, energy), this can present a great risk to operations. By outsourcing data management operations, an enterprise can diversify risk (using multiple sites or multiple vendors).
B) On-site Data Storage. Managing operations internally as opposed to outsourcing presents many of the same benefits along business functions. Data storage is no exception. Data maintenance requires servers and trained technicians – if an enterprise can afford and builds this option into the business plan, it can realize benefits in security, responsiveness, and control / customization. First, by maintaining on-site by company employees, enterprises can limit exposure of proprietary information. Second, in-house solutions are usually more responsive – though not always – given an enterprise’s inherent right to re-allocate resources (incl. personnel) within the business. Owning data management resources also provides the opportunity for the firm to customize its programming and processes to the business in a way which many subcontractors which not do, or would charge high fees for.
C) Private Cloud. Many of the benefits of the private cloud mirror those of on-site storage, but there are several additional benefits. The primary additional benefit of the private cloud is that has the potential to vastly increase efficiency of operations, as processes and personnel can be consolidated and managed centrally. While a distributed system benefits such as lower risk of disruption, by centralizing data storage solutions by moving data to an internally-managed cloud, enterprises can realize efficiencies of scale.
D) Hybrid Computing. Combining some of each of these three pure methods, hybrid computing can have several additional benefits. The one benefit which most interested my is reduced latency; local data storage combined with cloud enablers can reduce time for information processing. For me, it would be important to design an intelligent system which would make best use of local and outsourced resouces to reduce latency – for example, data is stored locally but processed centrally due to the strength of the computing systems at the centralized location. Of course this depends on how much bandwidth a firm has to transmit information. Another opposite example would be storing data centrally but processing locally; this is the inverse and the benefits would depend on the enterprise’s operating model. Both are valid options in today’s hybrid computing environment.
Risks / Disadvantages of Each Approach.
A) Public Cloud. The obvious risk for public data storage is reduced security. By providing data to a third party to manage, enterprises with high security requirements naturally incur some risk. Enterprises must trust their contractor’s security processes and their employees.
B) On-site Data Storage. By building and maintaining their data storage capabilities in-house, enterprises naturally incur additional equipment and personnel costs. All responsibility for building, maintaining, supplying (i.e., energy and other physical requirements such as the servers themselves, and wiring), and trouble-shooting the system – including the technicians themselves – is carried by the enterprise’s IT department. This responsibility transitions to the next possible risk – less flexible systems. As requirements change, grow, or shrink, a company with more physical infrastructure is less flexible because that infrastructure is semi-permanent. Owning and maintaining equipment also carries a capital depreciation cost, which is prominent in the technology sector where the life of systems is often shorter than that in other industries.
B) Private Cloud. Again the private cloud mirrors many of the risks of on-site storage, but differs primarily in its geographic approach. A private cloud offers a higher risk of data transit disruption – i.e., power outages or disruptions to networks. It is also less secure than the on-site storage, but more so than the public cloud. The private cloud may also present additional cybersecurity or physical redundancy risk due to its centralized approach – a single centralized system can often be more vulnerable to physical or cyber attack threats than multiple on-site, distributed systems.
Summary. The hybrid cloud approach for data storage can use any combination of the three components listed above, depending on factors such as the size of the organization, its products, and its general IT structure. Overall, using a hybrid approach decreases risks of disruption because maintaining multiple options to solve problems is always preferable to having a single option which could fail. Enterprise architecture and other business-IT enabling disciplines (IT infrastructure, business process development) can help enterprises efficiency analyze and select the right combination of these options to maximize value.
Sources:
1) Google Cloud. https://cloud.google.com/learn/what-is-hybrid-cloud#:~:text=A%20hybrid%20cloud%20is%20one,entirely%20on%20the%20public%20cloud.
2) Microsoft Azure. https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-hybrid-cloud-computing/
3) NetApp. https://www.netapp.com/hybrid-cloud/what-is-hybrid-cloud/