Summary and Key Takeaway
Whether by design or by chance, “hybrid IT” is becoming ubiquitous. But, opinions about hybrid vary among enterprise IT executives, providers, and industry pundits. Some consider a hybrid IT infrastructure to be ideal. Others consider hybrid IT to be a step along the path to a future consisting of only public cloud-based infrastructure. Our work with enterprises suggests three reasons why a hybrid IT infrastructure is – and, will remain – popular: project management, workload characteristics, and regulatory compliance.
The term “hybrid IT” or “hybrid cloud” can have various meanings. At ISG we accept the broadest meaning: a combination of two or more forms of IT Infrastructure ranging from traditional in-house, to private cloud, to private cloud. This is a broad spectrum spanning from:
- Light cloud usage. All mission critical core application workloads running on traditional in-house IT infrastructure and some non-critical, isolated applications running on a cloud-based infrastructure; to
- Complex cloud usage. Some mission critical core application workloads running on in-house IT resources and others running on one or more forms of cloud-based infrastructure.
During the early days of cloud-based offerings, some trade press and industry pundits were quite bullish and professed that “cloud-only” infrastructures or even more specifically “public cloud-only” infrastructures should be the universal objective and would become the de facto standard for enterprise IT within a few years. Somewhat perversely, this analyst maintained that cloud-based infrastructures are not the best alternatives in all situations. Rather, our analyses show that factors such as the technical and operational characteristics of workloads should be considered when selecting IT infrastructure.
Since those early days ISG Insights™ surveys and ongoing discussions with enterprise IT and provider executives reveal enterprises are increasingly combining one or more cloud-based offerings with existing in-house resources. More importantly, those discussions also indicate that for many enterprises such hybrid IT infrastructures are likely to remain for the several years…or for the foreseeable future.
However, some industry analysts persist in warning against hybrid IT infrastructures. Some claim that hybrid infrastructures are inherently inferior to an all-cloud infrastructure and some even label hybrid IT infrastructures as the “worst of both worlds” due to the following:
- Continued expense. The ongoing CAPEX and OPEX of the in-house site(s) and the associated IT infrastructure.
- Increased challenges. The exposures and challenges of managing and securing your data and processes with cloud providers on the external infrastructure. This includes establishing skills and processes that optimize both in-house and cloud-based operations.
Their view is that the result is not attaining the perceived benefits from cloud such as enabling staff to deliver higher value by working on innovative new solutions. Conversely, there are solid reasons for some enterprises to retain portions of their in-house IT infrastructure and migrate selected workloads to cloud-based alternatives. These include the following:
- Software Licensing. Some applications do not have license options that encompass cloud-based infrastructures. Upgrading or replacing the software can be challenging and/or costly.
- Sunk Investment. Existing in-house data centers represent sizeable investments in facilities and IT resources. Depreciation of those assets typically spans years. Writing off those assets early may have undesirable impact on the enterprise’s financials.
Thus, for some enterprise IT executives the question remains: In the evolution of IT infrastructure, is a hybrid IT infrastructure the desired endpoint or merely a waypoint along the journey to an all-cloud infrastructure?
Impact & Guidance
Managing a hybrid IT infrastructure is not trivial. The skill sets for managing an in-house data center (e.g., systems management) do not “translate” readily into the skill sets for managing a cloud-based infrastructure (e.g., contract management). However, our assessment is that despite the challenges associated with a hybrid IT infrastructure, we recommend that all enterprises should carefully consider the benefits of hybrid for their specific situations.
In ongoing discussions with provider and enterprise IT executives, we have identified three broad groups of factors that impact the efforts associated with modernizing or transforming enterprise IT infrastructure. Thus, to help in planning a path for infrastructure modernization or total transformation, enterprise IT management should consider the following:
- Project and Change Management. This group of factors includes all the aspects of managing and executing the modernization or transformation. Consider that, as mentioned above, adoption of a cloud-based offering (e.g., public cloud, private cloud, Software-as-a-Service solution) encompasses technical challenges, financial planning challenges, and people skills/management challenges.
- Thus, project and change management must reach beyond technical tasks such as installing software maintenance updates. Rather, change management must include training or hiring skills required to execute the project and to manage the ongoing operation of the “future state” environment. It is important to note that some of those “future state” skills will likely include expertise (e.g., problem diagnosis and resolution for cloud-based workloads) and skills (e.g., cloud provider contract management) that have not previously been needed within the IT organization. At the same time, staffing for some roles within the IT organization will be reduced or possibly eliminated.
- Workload Characteristics and Enhancements. These factors include the technical and operational characteristics for all application workloads. Each application workload should be evaluated to determine the infrastructure that provides the optimal platform based on the desired objectives of the modernization or transformation. For example, if the dominant objective is to maximize agility, public cloud will often be the best choice. However, if the dominant objective is to minimize the cost of ongoing operation, a traditional in-house infrastructure may be the best choice for some workloads.
- Another consideration is the location and the amount of data which is input to or output from an application workload. Large amounts of data which must be passed between applications on different infrastructures can easily cause significant charges for network capacity.
- Additionally, for each application workload being considered for migration to a cloud-based infrastructure, decisions must be made to select the amount of enhancement to be performed on the workload. For example, will the application be simply tested for compatibility and “ported” to the cloud with no or minimal enhancement; or will the application be enhanced to exploit cloud functionality such as dynamic resource provisioning.
- Certification and Compliance. This group of factors includes the actions required to attain or maintain certifications required by the industry(ies) within which the enterprise operates (e.g., insurance and health care) or by the government jurisdiction(s) in which the enterprise operates and its workloads and data reside (e.g., U.S., European Union). These factors must be thoroughly understood since they will influence selection not only of cloud providers, but also of cloud-based offerings (e.g., public versus private) and functionality (e.g., data replication, backup).
Not surprisingly, the challenges posed by the above groups of factors vary across enterprises and environments. In addition, our discussions with enterprise executives reveal that the relative importance of the above groups of factors varies across enterprises, across business units within an enterprise, and even across the inventory of application workloads within an enterprise business unit.
The groups of factors described above appear daunting. Some IT organizations opt for a simplistic approach such as migrating all workloads to a single cloud-based infrastructure. While this approach avoids the complexity of managing a hybrid environment, it is likely to result in unnecessary costs for operation of some workloads.
Our research shows that the optimal IT environment is attained through a process of selecting the optimal infrastructure for each workload. This entails all three groups of factors and typically results in a hybrid IT infrastructure. In addition, IT management should recognize that pricing for cloud-based IT resources is not static. Similarly, salaries, utility costs, etc. are not static. Thus, IT management should commit to an ongoing process of re-evaluation of workloads to ensure ongoing optimization of the IT infrastructure.