Introduction
Hybrid cloud is a type of cloud environment where different clouds and/or on-premise infrastructure function together, as interconnected mechanism. The benefits of hybrid cloud are that it is more variable, allows for faster application development, and combines the better qualities of other clouds while mitigating their disadvantages. For example, native cloud natively utilizes services and infrastructure from cloud computing providers such as Amazon. Cloud native technology, alternatively, gives an organization the possibility to create their own applications for public, private, of hybrid clouds. On-premise means that the servers where the organizations keep their data are local and stored in the building where the organization is based. Thus, the possible models of an organization’s infrastructure are considerably diverse.
Naturally, native clouds have many advantages, such as the ability to stay online during energy outages, and, as mentioned above in relation to hybrid cloud, update their applications. Their applications are smaller and more mobile and simple to develop as well. Cloud native applications use DevOps automations features and enable continuous delivery and deployment of software changes that get released on a regular basis. However, they become more expensive with time and require a stable internet connection, while their applications may be more effective, but are less configurable.
As there are many tasks that organizations face, each of them being more or less easy to deal with through different infrastructure, hybrid cloud gives an optimal option of using each cloud’s advantage in each separate case. Organization may prefer to move some of their data to a public cloud while retaining some of the more sensible information on-premise. Thus, the reason to migrate to a hybrid cloud is generally the aim for versatility.
Literature Review
Kataoka, B., Brennan, J., & Aggarwal, A. (2022). Digital Transformation and Modernization with IBM API Connect. Van Haren Publishing.
The authors present the purpose of the book as to share API Connect experience and implementing hybrid cloud for it, thus stating their implementation of the hybrid cloud. According to their guide, API Connect can be installed on Azure, GCP, Amazon Web Services (AWS), and IBM Cloud. The researchers note that the benefit of this is no hindrance by preexisting infrastructure and outdated legacy applications. There is an opportunity to mix and match the deployment in accordance with the digital modernization as well. A successful hybrid cloud needs a common platform and the ability to integrate with new cloud applications and on-premise systems. Lastly, during the move towards a hybrid cloud, it is important to keep flexibility intact. The on-premises functionality includes applications utilizing SOA, Java, messaging, and web services. To conclude, the authors deem hybrid cloud to be an optimal way to use API Connect but warn that hybrid clouds have a lot demands to function properly after its implementation.
Zikopoulos, P., Bienko, C., Backer, C., Konarski, C., & Vennam, S. (2021). Cloud Without Compromise: hybrid cloud for the Enterprise (1st ed.). O’Reilly Media.
The authors define hybrid cloud as a unification of public cloud, private cloud, on-premise infrastructure, and management, creating a single, flexible, cost-optimal IT infrastructure. Companies can combine different cloud services to create their optimal combination to attend to their technical and business objectives. Developing cloud native allows developers to make monolithic applications into business-focused functionality that can be run and reused anywhere. It assists in modernizing applications faster and connect cloud services to data on cloud or on-premises infrastructure as well.
Hybrid clouds are a necessity, as without them, customers may be forced to run on-premises, legacy applications that they cannot relocate to the cloud. However, as the authors admit, hybrid clouds, being more complex, are difficult to work with. For developers, they mean multiple environments to code and test. For operations teams, they lead to more chances of failure, more late night shifts, and more difficult integration. For security teams, hybrid cloud means a larger space susceptible to attacks and complex security policies.
The authors’ example of the usage of the hybrid cloud is the social-mobile-cloud. It has accelerated social change and changed the global flow of information. Social-mobile cloud can limit the spread mass media, shifting attention via a many-to-many model. Another usage of hybrid cloud is in automation. Summarizing, a virtual private cloud allows developers to have more ways that they work in through the “everything as code” and “as-a-service” delivery model of hybrid multicloud vendors.
Shrivastwa, A. (2018). Hybrid cloud for Architects. Van Haren Publishing.
The author admits that the word cloud has been commonplace in the industry and marketplace for over a decade. The definition he considers the most widely accepted is that the hybrid cloud is an environment comprising of a Private Cloud component (on-premise) and a Public Cloud Component (third party). Discussing advantages and disadvantages, the author states that different clouds are better suitable to different needs. For example, private clouds are better for work with an always-on virtual machine, especially when there is sensitive data on it. Public clouds are better for work with big data, development and testing, as well as cognitive services adoption and rapid prototyping. Self-healing applications are suitable for both types of cloud.
The author names two major services required to be set up beforehand in order to use a hybrid cloud: network connectivity and DNS service. As for his supposed usage of the hybrid cloud, he lists the Supporting Applications Use Case, the Isolated/Distributed Use Case with the help of a Cloud Management Platform (CMP), and cloud bursting which augments the capacity in the private cloud/private data center. Thus, the book is objective when discussing hybrid clouds, admitting both its advantages and disadvantages.
Dutt, D. G. (2019). Cloud Native Data Center Networking: Architecture, Protocols, and Tools (1st ed.). O’Reilly Media.
The book mostly deals with native cloud, as per its name. It focuses on connectivity options and on the work to be done at the on-premises data center. Still, the author discusses hybrid cloud as well, defining it as dividing a data center between the enterprise’s on-premises servers and a cloud provider’s servers, with them communicating with each other transparently. As for the usage, Dutt states that many enterprises use the hybrid cloud model as a way to move certain parts of the business to the cloud. According to Dutt, the most common model for running a service inside a public cloud is a Virtual Private Cloud (VPC) instance, which consists of multiple compute nodes connected via an L3 network. Another notable part of the author’s work is the study of the ways in which it is possible to use hybrid cloud to provide seamless routed connectivity between on-premises data center and the VPC instances in the cloud. Thus, while Dutt’s work is dedicated to native cloud, it still deals with some of hybrid cloud’s functions as a means to achieve certain needs.
Reznik, P., Dobson, J., & Gienow, M. (2019). Cloud Native Transformation: Practical Patterns for Innovation (1st ed.). O’Reilly Media.
This is another book dedicated to native cloud and discussing hybrid cloud as well. According to the authors, hybrid cloud mixes on-premises infrastructure with private or public cloud services best suited to the company’s specific needs and usage. The most important of the authors’ mentions of disadvantages is their list of reasons for the failure of cloud native transformation. Namely, they list difficulties due to the complexity of distributed systems, the relative immaturity of the cloud native ecosystem, and the failure to adapt and evolve organizational culture to keep pace with changing technologies and delivery expectations.
Another statement of disadvantage, as well as an example of usage, is a hypothetical company with Waterfall or Agile culture, which has attempted a “lift and shift” strategy to move their current system onto the hybrid cloud. The authors argue against the usage of hybrid cloud in this case, stating that operating costs will significantly increase. The authors advise to improve the organization’s culture instead, preparing it to the shift in the future while staying on-premise for the time being. This is book is therefore similar to Digital Transformation and Modernization with IBM API Connect in that it warns of the necessary preparations to implementation of hybrid cloud.
Chudasama, V., & Bhavsar, M. (2020). A Dynamic Prediction for Elastic Resource Allocation in hybrid cloud Environment. Scalable Computing: Practice and Experience, 21(4), 661–672.
For the authors, a hybrid cloud is an affordable solution to deal with a burst in requirements for specific events. In such a case, an application is running on-premise and bursts to a public cloud for more resources. The stated advantages of this type of hybrid cloud are cost reduction and scalability with data sensitivity, increasing user satisfaction. In order to achieve such a versatility of the services and promised Quality of Service, the authors state the need for resource management in Cloud Data centers. The main goal is thus to provide services with a dynamic policy which adds or removes storage and computing resources to enhance the application performance. Cloud providers such as AWS or Google provide flexible features on some metrics (CPU utilization, memory). Thus, the article’s utilization of hybrid cloud is for resource allocation is specific burst events, with the authors leaning into a Bi-directional LSTM based approach.
Park, J., Kim, U., Yun, D., & Yeom, K. (2020). Approach for Selecting and Integrating Cloud Services to Construct hybrid cloud. Journal of Grid Computing, 18(3), 441–469.
The authors define a hybrid cloud as a cloud deployment model that combines a public and a private cloud, or a public cloud with an existing on-premise environment. They state that this is done to achieve cost-effectiveness and security by storing security-sensitive data in a private cloud or on-premise environment, while storing less personal or secret data in a public cloud. Thus, the definition and the idea of hybrid cloud’s functions is relatively standard.
The authors admit some disadvantages of creating a hybrid cloud and trying to determine an optimized cloud strategy. Their reasoning is that the architecture of the hybrid cloud has to be constructed by appropriately combining private and public clouds, used as a unified cloud environment. With not every company or organization currently having cloud experts, they end up employing inadequate solutions to try to transition to a hybrid cloud environment. Despite some public cloud companies providing consulting services, they are limited, and, in turn, limit the resulting hybrid clouds to their own traditionally provided variants. This not only limits the flexibility of the hybrid cloud, but makes impossible to fix problems caused by the vendor’s standard model as well. The authors therefore emphasize that it is vital to have competent cloud experts before attempting to implement a hybrid cloud.
Rong, H., Liu, J., Wu, W., Hao, J., Wang, H., & Xian, M. (2020). Toward fault-tolerant and secure frequent itemset mining outsourcing in hybrid cloud environment. Computers & Security, 98, 101969.
The authors of the article define a hybrid cloud as a setting, in which the public unreliable cloud and semi-trusted cloud cooperate to mine frequent patterns over the encrypted database. In the authors’ system model, they mainly focus on how hybrid cloud servers process frequent itemset mining request while preserving privacy. They have their own terminology of non-hybrid clouds, with the cheap unreliable cloud (UC) composed of a large number of unreliable servers, providing public storage and computation services. The Semi-trusted Cloud (SC) is more expensive, but provides better service security and integrity It comprises a semi-honest server coming from private cloud or commercial public clouds. This better reliability of the SC is possible due to strict Service Level Agreements.
UC servers store the encrypted outsourced datasets, all the intermediate results and final outputs, while relying on a SC to execute the mining protocol. The SC plays the role of clients’ proxy, and assists the UC in performing auxiliary calculations. When working together, the UC and the SC constitute the hybrid cloud environment. With UC servers holding the shares of the joint database and the support threshold, the hybrid clouds collaborate to find all the frequent itemsets. The protocol protects privacy of the out-sourced database and itemset supports under the hybrid-cloud threat model. Thus, while a UC is an inferior model, its disadvantages are mitigated when functioning in a hybrid cloud environment, and it still has its role in that case, which is an advantage of hybrid cloud.
Aryotejo, G., Kristiyanto, D. Y., & Mufadhol. (2018). Hybrid cloud: bridging of private and public cloud computing. Journal of Physics: Conference Series, 1025, 012091.
The authors define hybrid cloud model simply as a combination of private cloud and internet. The most valuable part of the article is the comparison of four cloud deployment models, namely Private, Public, Community and hybrid cloud, listing the advantages and disadvantages of each. Private Cloud, which belongs to a single private organization, has higher security that others, but its cost is high. Community Cloud belongs to two or more private organizations with identical requirements, its security level is lower, and the cost is medium. Public Cloud belongs to a Cloud Service Provider (CSP), its security is low, and its cost is taken per usage. Hybrid cloud belongs to CSP and private organizations, its security is lower than Private and Community Cloud but higher than Public Cloud, and its cost is taken per usage as well. The main usage of hybrid cloud the authors suggest is to solve the security issues problems Public Clouds have. This is possible because hybrid cloud as a server provisioning system simplifies local-based cloud computing application via online. Thus, for Aryotejo et al., hybrid cloud is a troubleshooter for Public Cloud’s unreliable security.
Jain, T., & Hazra, J. (2019). Hybrid cloud Computing Investment Strategies. Production and Operations Management, 28(5), 1272–1284.
To define a hybrid cloud, the authors refer to the National Institute of Standards and Technology (US Department of Commerce) instead of doing it themselves. The Institute defines a hybrid cloud as a “composition of two or more distinct cloud infrastructures (…) that (…) are bound together by standardized or proprietary technology that enables data and application portability”. The authors state that implementing a hybrid cloud is the optimal strategy in a situation with a medium average demand. In hybrid cloud strategy, an establishment can manage its resources on-premise while having other resources provided externally by public cloud. This may become a necessity when the establishment cannot meet its entire demand based on on-premise capacity only. A hybrid cloud computing gives the option to outsource the excess demand to an external public cloud. As a result, projects can be completed in less time with lower risk and lower administrative overhead. Thus, the IT unit gains an opportunity to ensure rapid provisioning and integration of new services. Other advantages of a hybrid cloud the authors list are unlimited storage space, improved scalability and flexibility, and investment savings.
Methodology
The first step of moving an app to a cloud is deciding what to move, namely, what specific applications, workloads, and data the organization will be transferred to the cloud, and which ones should not be moved, as well as the quality, scope, and volume of the transferred data. Then, the technology trade-offs will be addressed, including the selecting of the best approach to migrate the application and deal with important issues such as security, governance, and disaster recovery. Other factors are cloud bursting capabilities, time constraints, and possible lag.
As its next step, the company will define the requirements of its core Enterprise resource planning (ERP) system. Every ERP system, even in a hybrid deployment, needs a core that includes functions like logistics, finance and procurement. ERP system is critical to the success as a driver of enterprise-wide digitization, being the tool for recording and managing of all the critical data for material, employees, processes, and products. Thus, a cloud is useful for providing access to the central ERP system at all times, but they must be made compatible.
The next step is the app preparation, including configuring virtual machines and performance settings. They do not necessarily have to be converted into new formats, as some APIs can work across various cloud platforms and enable the organization’s data and applications to function across different environments and platforms. Still, all the data and applications must be stress-tested in the new environment and optimized to ensure optimal performance. It is important to build a diagram of dependencies between applications as well, as, if moved improperly, individual applications may cause business risks disrupting the usual performance of current tasks. Therefore, the application has to be ready for the transfer, the functions and dependencies of each of its parts must be established to exclude the possible problems caused by the application itself. This is especially important considering that the procedures planned to be executed further are complex by themselves, and do not need further complications by the improper configuration of the application.
When assigning hybrid cloud agreements in a multi-cloud environment, options that interface with other platforms will be assessed. When choosing a hybrid cloud framework, it must be taken into consideration that there are no clouds the organization currently uses. The reason for it is that some hybrid cloud frameworks have certain compatibility attributes which are influenced by an absence or a presence of a public cloud. As the transfer to a hybrid model is being done primarily to minimize costs, the Eucalyptus solution will be utilized. To complete the transfer, AWS DataSync will be used, as the Eucalyptus only works with data in an AWS format.
Then, a Data Migration Plan will be created, the main aspect of which is to determine whether the cloud experts of the organization have the necessary skillset to manage the transfer of data to hybrid cloud. If this is not the case, the services of a cloud-migration service provider will be utilized. This is not to be taken lightly, as an incomplete understanding of data by the cloud expert may lead to code errors, extensive data remodeling, and extra work from transferring obsolete or duplicate data to the hybrid cloud. With the hybrid architecture and support strategy fully planned, the migration of applications and data from the public cloud into the hybrid environment can begin.
With all the preparations completed, the next step will begin. It will start with making an informed decision as to what changes each application requires and what approach will work best when rehosting on a hybrid cloud. The application will be modified to use application-level security services, such as the entity-based ones. This will require modifications to the application to utilize this type of security, such as using Security Assertion Markup Language (SAML), the open standard for federated identity. For rehosting, there are three possible options: to lift and shift, partition, or refactor the application. As the cloud experts have the necessary qualifications, the refactoring approach will be taken. Despite it being costly, while the organization obviously prefers to save money, it will guarantee that the moved application will function smoothly and beneficially. Refactoring the application usually includes a complete rewrite of it to take use the hybrid cloud to its full potential. However, as mentioned above, this approach is costly, with refactoring demanding redesigning and rewriting the application anew. A development team will have to work for several months or more on the rework of the application.
When ready, the components of the application will spread between the public and private cloud, with refactoring breaking it apart precisely. The division of the parts will be done in accordance with the possible maximization of performance while keeping operating costs as low as possible. Then, AWS DataSync will be used to ensure that the refactored application will function on Eucalyptus properly. The application’s functionality will then be maintained by both the organization’s cloud experts and AWS employees. This methodology, despite being costly, is the only means to ensure that the application will work smoothly and beneficial.
Conclusion
The conducted research brought understanding of the nature of hybrid cloud, its advantages and disadvantages, and the dangers of migration to it. Namely, hybrid cloud is an environment where a public and a private cloud function together, dividing the data of an organization between each other. The obtained skill is the theoretical ability to decide upon the strategy needed to successfully migrate an app or any data to a hybrid cloud.
Hybrid clouds are much more variable than other types of cloud such as purely public of purely private. They allow for a faster modernization of applications, while users with the on-premise model or a private cloud have to use the outdated legacy applications. The other advantage of hybrid cloud is that it has better security than a public cloud, while still utilizing some of its functionality, which is uncommon for private clouds. Hybrid cloud is useful during bursts in demand characteristic to specific events as well. Its more focused uses are API Connect or a social-mobile-cloud.
Still, the hybrid cloud has a number of disadvantages. For example, it may be expensive as it is usually paid for per usage, and it is significantly demanding on the skills of the cloud experts an organization employs. If the migration is being done from a different cloud, such as a public or a private one, the organization may face compatibility issues between them. To make sure that the migration ends up successful, it is possible to hire the designer of the specific hybrid cloud being used. In that case, despite the successful migration, the organization may face specific troubles common for the product of the cloud designer, which will be impossible to fix. Thus, the optimal strategy is to employ qualified cloud expert at the organization.
References
Aryotejo, G., Kristiyanto, D. Y., & Mufadhol. (2018). Hybrid cloud: Bridging of private and public cloud computing. Journal of Physics: Conference Series, 1025, 012091.
Chudasama, V., & Bhavsar, M. (2020). A Dynamic prediction for elastic resource allocation in hybrid cloud environment. Scalable Computing: Practice and Experience, 21(4), 661–672.
Dutt, D. G. (2019). Cloud native data center networking: Architecture, protocols, and tools (1st ed.). O’Reilly Media.
Jain, T., & Hazra, J. (2019). Hybrid cloud computing investment strategies. Production and Operations Management, 28(5), 1272–1284.
Kataoka, B., Brennan, J., & Aggarwal, A. (2022). Digital transformation and modernization with IBM API connect. Van Haren Publishing.
Park, J., Kim, U., Yun, D., & Yeom, K. (2020). Approach for selecting and integrating cloud services to construct hybrid cloud. Journal of Grid Computing, 18(3), 441–469.
Reznik, P., Dobson, J., & Gienow, M. (2019). Cloud native transformation: Practical patterns for innovation (1st ed.). O’Reilly Media.
Rong, H., Liu, J., Wu, W., Hao, J., Wang, H., & Xian, M. (2020). Toward fault-tolerant and secure frequent itemset mining outsourcing in hybrid cloud environment. Computers & Security, 98, 101969.
Shrivastwa, A. (2018). Hybrid cloud for architects. Van Haren Publishing.
Zikopoulos, P., Bienko, C., Backer, C., Konarski, C., & Vennam, S. (2021). Cloud without compromise: Hybrid cloud for the enterprise (1st ed.). O’Reilly Media.