2017年12月5日 星期二

Problems to be considered before the of an enterprise data center migration

In the data center migration: the development trend of the digital transformation of the industry leading a trend which cannot be halted, there are two kinds of enterprises are faced with a clear choice: or the implementation of modern transformation and business migration, or missing things, artificial intelligence and machine learning, AR / VR, cloud computing, big data analysis, mobile technology development opportunities. Which business may therefore be interrupted or replacement.
 Data centre migration
The pressure of business transformation and modernization depends largely on the IT team. They are the gatekeepers of data centers, so they are also the key driving force for future opportunities and success of enterprises. As the modern data center infrastructure becomes "cloudy", its planning is crucial. When the era of modernization is coming, most enterprises are faced with their own internal implementation, or the difficult decision to cooperate with the outside third parties.
A new study commissioned by DELL EMC company Forrester has found that enterprises that realize data center modernization and business migration cost more than 4 million dollars on average than those from cloud computing providers. And these modernized enterprises take two years to complete the process and transfer the responsibilities of their IT staff to their aspects.
However, if the enterprise is planning to implement the process and migration of the modernization of the data center, it is important to have realistic expectations and clear strategies. The enterprise can consider the following aspects carefully:
(1) enterprise expectations
Although the data center infrastructure or operation team of an enterprise has its own thoughts on the speed of modernization and migration, these assumptions are often not verified by other stakeholders in the enterprise.
In order to ensure the consensus of team members' ideas and awareness, the team should consider the following questions:
What stakeholders do companies need to communicate with? What information does everyone in the team need? Companies need to identify those who need to participate, and ensure their support for modernization and migration.
Who is responsible for project coordination and communication in an enterprise team? The person in charge needs the skills of the organization and managers to ensure that the plan is in place, and all the key stakeholders need to work together.
What is the tolerance for downtime? If the IT team estimates business migration will lead to downtime 12 hours, and in charge of the enterprise can only tolerate 4 hours of downtime, so enterprises need to reconsider the migration plan.
(2) expectation of resources
This transformation needs to be seen as a long-term process, and data center migration is a complex event that most IT departments rarely perform. The results of their execution will vary greatly, depending on whether the enterprise is to choose to do it yourself or to seek the help of the supplier.
According to the research of Forrester company, if the company chooses to cooperate with the supplier, it can shorten its implementation time by 40%. The migration will be completed earlier, and the enterprise will run the data center at less actual cost. Enterprises also need to consider which resources can be invested in this change.
Before data center migration starts, enterprises need to map the dependencies between all applications, servers and infrastructures, so as to ensure that they check all resources and create clear strategies.
(3) schedule
Large enterprises may have multiple enterprise level data centers, which means that thousands of servers and hundreds of applications are keeping their business running. However, there is only a short downtime in the implementation of migration. Stakeholders of the business need to tolerate this, and businesses may not have enough time to complete the complete migration of all tasks at once.
Instead, the migration plan should divide the process into several stages, and divide the work according to the granularity of application environment, the granularity of data or information, or the order of application upgrading. For example, an enterprise can give priority to an upgraded application or system.
(4) from the A point to the B point
Of course, the modernization and migration of data center projects may require the purchase and replacement of new hardware, and the need to build new data center facilities. Enterprises need to ensure that the following points are considered:
Have you considered the environment needed in the new facility?
When can the enterprise's new facilities and environment be enabled?
Do businesses need temporary equipment to migrate applications and environments in order to use new devices in the future?
If enterprises migrate their business from locally deployed data centers to the cloud, or migrate between clouds, or have a cloud infrastructure, another layer of risk is also needed to consider. Compared with the data center environment deployed locally, IT teams usually have low visibility to public cloud applications and resources, which makes migration planning and execution more difficult. It is essential to assess the process of meeting the cloud applicability of business and technical requirements.
(5) supplier commitment
Since the modern data center not only involves updated hardware, it is necessary to include third party software providers into the workflow from the very beginning, and we need to ask the following questions.
If an enterprise has problems in the process of migration, is the supplier willing to provide support?
What are the suggestions for vendors to upgrade the software before or after the migration of the data center?
When the migration is implemented, are the products or services supplied by the third party available?
Enterprises need to ensure that their suppliers provide a support plan to ensure that there is a reliable strategy to solve potential problems before migration.

沒有留言:

張貼留言