Data Migration - Term Paper Example

Data migration is the process by which we perform data transfer from one data storage system to another, from one data format to another or between different computer systems. Companies are increasingly aware of the advantages of migrating their servers to the cloud. The cloud ('cloud') has become a paradigm that offers optimization of operating costs, reduces risks, offering the possibility of a lower initial investment, provides access flexibility and the efficient use of resources. The offer of clouds in the digital world is wide, and in most cases, it can be difficult to decide what type is ideal for a companys entity.

Implementation of the Cloud Server Migration of UIS

Cloud computing, also called cloud computing, refers to the regime of outsourcing, in which third-party services are contracted using agreements of the level of service ("Service Level Agreement," SLA) and through internet protocols. This is possible because several companies have chosen to optimize the use of its technological infrastructure, offering storage solutions ("hosting") and computer services ("outsourcing"), through subscription policies and pay-per-use, among others (Khadka, Saeidi, Jansen, Sage & Haas, 2013). The migration of legacy systems from an organization to Cloud computing environments represent large cost benefits (pay only for what is used, less investment in hardware and maintenance, etc.) and the opportunity to make changes in the business model to adapt quickly to demand of the market and its dynamic environment. In the same way, the migrations of systems offer the possibility of integrating, in the same computer solutions, the applications of the organization and create collaborative processes with clients, partners, and different suppliers.

Migration preparation and business planning

Most of the works analyzed in the area of software engineering focus in the functional aspects to decide on the migration towards services of cloud computing, for example, the work of the migration as an adaptation of the applications of legacy systems, to be executed in the infrastructure of cloud computing providers. There are four types of migration: component replacement, partial migration of the application, total migration of the application, and transform the application to cloud computing (Kwon & Tilevich, 2012). At work migration processes are also left aside, and the aspects of acquisition, implementation, economic factors, security, and privacy, of a classification of four types of migration: migration of data, information migration, services migration, and autonomous migration. Undoubtedly, the trend is to transform traditional applications first, commonly structured and forms-based, to applications based on the service-oriented architecture ("Software Oriented Architecture," SOA).

This transformation favors weak coupling, reuse, integration and the abstraction of the technological infrastructure, which are critical factors in the environments of cloud computing. The process presented in the cloud computing consists of the evaluation of the scalability of the components, evaluation of the orchestration, identification of the components to be refactored and the evaluation of the solution in front of the destination cloud computing environment.

Next, it presents the characteristics of cloud computing and the different types of migration of legacy systems, according to the analysis of the features of these systems and the models of cloud computing implementation.

In the majority of the occasions, the companies look for a solution to: create infrastructures, develop modern applications, obtain information from data or manage identities and accesses. To do this, before deciding, very often, they issue an assessment among the following contenders: Amazon Web Services (AWS), Google's Compute Engine (GCE), HP, IBM (SoftLayer), Microsoft Azure and Rackspace (Zhang, Cheng & Boutaba, 2010). It is necessary to carry out a reflection of the migration process and its implications for the particular company through the following steps:

Analyze internal processes and assess business needs.

What will be the exact use of the cloud?

What applications or workloads do you want to take to the cloud?

What are your security requirements? Private or public cloud?

Study how the new technology will adapt to your business

What are the implications for your current IT team?

Are your licenses and software compatible with the cloud?

In what state are the data you must migrate to the cloud?

Determine the conditions of partnership with the provider

How will the interaction with the provider be?

How long will it take to migrate?

How will the plans and prices be established?

What type of support is offered?

What will be the level of portability to change the environment?

Server migrations are essentially tool-based and involve the migration of complete servers (physical or virtual) to private or public clouds. The idea is that the application (s) of the server continues working, as they are, after the migration. These transfer operations do not imply correcting code and, therefore, are faster and cheaper, and involve a comparatively lower level of risk (Winkler, 2011). However, there is no margin to modernize these applications or integrate them with native database platform services, middleware, or PaaS, during the process. Normally, these server migrations work well, as long as the current versions of the application, the application servers and the runtimes on the destination platform are maintained.

Application migration is the preferred approach, when some level of code correction is required, which takes into account the variations between the source and destination environments. These modifications can be derived from different operating systems, application servers or execution times. Application-based migration is also preferable in scenarios where the client requires integration with native services, which maximize the benefits of the public cloud and minimize the TCO of private cloud applications (Cattedu & Hogben, 2009). As they usually require some code correction, application migrations are not as fast as server migrations. The risk and associated costs are also higher and depend on the degree of automation that is achieved. Application migrations are commonly used in scenarios involving UNIX to Linux migrations, Windows / Linux updates, application server updates (JBoss, etc.) and runtime updates (.Net, Java, etc.).

Windows Azure for UIS

One advantage of Windows Azure is its easy integration with other Microsoft tools, competitive packages and security guarantee with multi-cloud back-up. It is possibly the most important added value offered by Microsoft is the close relationship with its customers. For decades it has marketed its services on the corporate front allowing them to offer attractive service packages, confidence guarantee and ease of acceptance among employees.

In a survey by Gartner (a consulting and information technology research company based in Stamford, USA) 64% of Azure users surveyed said that the trigger for their decision was their previous relationship with Microsoft (Andrikopoulos, Binz, Leymann & Strauch, 2013). What's more, much of Microsoft's advantage lies in its ability to integrate its platform with other Microsoft products that are well established within companies in a global environment. For example, an IT team that has invested time and resources in Office 365 or the Hyper-V platform will surely consider Azure as a natural extension of its current operations.

Azure is the only platform in the cloud that Gartner has qualified as a leader in the sector for infrastructure as a service (IaaS) and platform as a service (PaaS). The effective combination of managed and unmanaged services allows you to create, deploy and manage applications to obtain high levels of productivity (Chauhan & Babar, 2011). Azure does not limit consumers to Windows products, it even offers Windows, Linux or BSD machines at prices that compete with Linux, includes most open source technologies as well as a variety of third-party proprietary tools and supports any operating system, language, tool and framework, be it Windows, Linux, SQL Server, Oracle, C # or Java.

There is the possibility of combining a data center and the public cloud, what Microsoft calls a hybrid. A passport to the best of both worlds, being able to expand the IT options without added complexity. Data storage, backup and recovery are efficient and economical. It is also easy to create applications that cover both the local infrastructure and the cloud. Azure can be expanded or reduced quickly to suit any demand, this is its popular 'pay per use'. There is a billing per minute option and Microsoft also commits to matching the prices of the competition for the known infrastructure services, such as process, storage and bandwidth. In addition all the information stored in Azure can be exported to other environments without any restriction.

The Latest, Containers

Another emerging approach in this type of migration is the so-called container strategy. The idea is to decompose an application or server into components, such as application servers, web servers, data services, etc., and then package each component separately, as a container (micro-VM). The containers are then sent to the customer's container of trust, for its implementation (Kaisler & Money, 2011). The containers offer the highest level of portability, regarding workloads, and the highest level of utilization of system resources. The effort to contain a workload is greater than that derived from a server migration but equal to or less than that of application migrations.

Modernization versus Migration

Once the applications are migrated to the cloud, either from the server or the application, the next logical step would be to modernize these applications to a native cloud. This would allow applications to take full advantage of the Cloud delivery model, integrating and automating Cloud services, with DevOps, microservices, and containers. The modernization exercise requires an evaluation of the application that determines if it needs a new architecture that addresses new frameworks and patterns. For legacy technology applications, a native Cloud modernization could also involve a technology conversion.

Migration to the UIS Company's Cloud

The easiest method to migrate is to recreate the website of the company, database or file server entirely on the new servers, configure and test them, try them again, and then change from old servers to new ones. Although this requires a lot of tests and qualified system administrators to configure their new servers, the migration strategy and time is relatively simple. Just make sure to freeze all development activities on the website or application while configuring the new server, and synchronize the data between your old and new database before changing the DNS of the web address.

In fact, the migration strategy can be complicated due to the need to provide normal services throughout the entire migration process. Although moving a series of files is not a complicated process, moving a fundamental business service or a live updatable database has several challenges. One of them is having to avoid downtime, maintain user access and permissions, and maintain the continuity of the data (Mell & Grance, 2011). Before making the migration, the company has to consider the time and impact of freezing a development. Changes to scripts, files and settings cannot be made. Freezing all changes during the migration process will ensure that the migrated files are the same as the files on the source server. The company should plan the time it will take to meet the require...