The data center migration, according to the survey agency, is expected to increase three times in the next five years. By 2020, Internet of things (IOT) devices will be popularized in a wider range. It is expected to reach 30 billion units, and the bandwidth of enterprise data center will be overburdened in the near future. The same is true of the same data center hosting facilities, and the growth of the enterprise's own construction data center is almost always exponential.
Needless to say, data center managers must face some very serious challenges: the need for higher speed and channel delay performance of ultra low application; increase support leaf and spine network port density; and the enterprise can find a way to improve the availability of the network method, at the same time reduce the cost.
For many data center managers, the answer is to migrate their infrastructure, either updating over time, or replacing them all in order to support new speed, delay and port density requirements. Is this necessary for all data center facilities? Not at all。
On the one hand, people cannot ignore the staggering current and future trends in data use, and will not be stable in the foreseeable future. Therefore, when the infrastructure migration is a problem that happens. On the other hand, each data center facility has some very unique business needs, stakeholder expectations and technical considerations.
According to reports from the industry media, Equinix has acquired Telecity Group on a global scale for $3 billion 600 million, and has obtained more than 40 data centers in Europe. At the same time, apple, Microsoft and Google are investing in data center facilities in Ireland, Holland and Scandinavia. Large suppliers and companies in the United States are strengthening ties with the Atlantic end-users, and IBM recently opened two cloud data centers in London. Therefore, the migration of infrastructure has become a matter of time. So, if not now, then when is the correct migration time?
Pay attention to the standard. One of the requirements for the formulation of new standards is to demonstrate the necessity and feasibility of the new standards. In other words, if there is not much market demand, organizations like IEEE, TIA, and ISO generally do not start the recommended standards. So if a standard is being developed, it can be determined that there is enough market demand in the standardized market.
What is the cost of sticking to the operation? If the enterprise decides to delay the migration data center, then the decision must be clear and clear. New or sudden changes in market dynamics may increase bandwidth demand overnight. With the growth of data center requirements, how does this impact the enterprise infrastructure, and how can it affect manageability and potential downtime? All these problems require careful consideration by the management of the data center.
The bottom line is that the enterprise's data center needs to play its specific role. It is faster to migrate, so Yan need to ask yourself this simple question: "the development of our physical layer does not interfere with the business?"
沒有留言:
張貼留言