The virtual datacenter will have the architecture "no Walls" data center

In the 2012, everyone looked at the use of the data center's internal operations to enable cloud computing. And when we step into the 2013, enterprise it needs to consider the network that connects the data center, especially the need to consider how these cross data center networks should be tuned to support new cloud computing use cases and associated network requirements (bandwidth scalability, low latency, security, virtualization, and automation).

There are several main reasons for this. First, the enterprise IT department is very active in improving efficiency. Businesses have become more efficient from their networks, first through data center consolidation, then virtualization. According to some studies, cloud computing will help businesses save up to 25% of their IT service infrastructure and hardware costs. At the same time, the network is also the main catalyst, cloud service providers can use the network to run multiple data centers as a shared pool of virtual data center. We found through our own research that this could help reduce the overall cloud data center resources by 35%.

Second, in many enterprises, the use of cloud computing services is still in its infancy, and cloud computing software and infrastructure, a service (IaaS) application, now located in the service Provider cloud computing environment, are often locked behind a semi proprietary architecture. As cloud computing deployments become more complex, it will need to deploy more open architectures while setting standards for APIs and management tools. By doing this, organizations will be able to combine private data centers with multiple vendors ' cloud computing data centers to achieve higher workload mobility and the ability to deploy new applications. And this will require a better network to run efficiently.

How do companies use cloud computing now?

At present, companies typically use cloud computing infrastructure to back up storage to cloud computing data centers through slow IP networks, moving data slowly because the network allows the north-south transmission stream from user to machine across multi-tier IP network architectures.

In the future, in the face of mass storage, companies may want to take full advantage of lower-cost cloud computing storage. Enterprises will need to have a network, that is, when the data needs to move, the network can dynamically respond to TB data movement, without bottlenecks, security vulnerabilities or discarded packets and so on. This will require a stream of things from machines to machines in the data center (and more and more data centers), which will have more stringent quality of service requirements.

In the near future, we will see applications such as storage synchronization, data center storage virtualization, and virtual machine migration driving more data workloads. This means that the cloud computing network needs to be designed in terms of performance to meet the changing needs of traffic and the high intensity of future challenges. For example, a more flat 0/1/2 layer network between data centers can provide better scalability, low latency, and more accurate performance. Through this architecture, even with the heaviest load, the network will not become a bottleneck, and it can provide the lowest cost per bit throughput.

Network is a key factor in the future use of cloud computing

As companies increasingly start using cloud computing, IT departments will need to find ways to automate cloud computing using intelligent networks. The Software Definition Network (SDN) is the key to achieving greater flexibility and lower costs. policy-based software automation tools will facilitate the orchestration of workloads between cloud computing data centers and between enterprises and cloud data centers. Requirements for performance parameters, such as bandwidth scalability, will be displayed through a high-level software interface that connects to the network control plane, rather than being implemented by the operator using the command line. This on-demand performance will be reflected in the application layer to ensure that the network and cloud computing adjustments meet enterprise requirements. To do this, the network needs to be an integral part of the entire cloud computing architecture, a key factor in all parts, providing flexible application and service delivery as a "backbone" across the data center.

Cross-datacenter networks also need to be virtualized. Virtualized networks can partition resources in many different ways, such as virtual circuits (EVPL), virtual wavelengths (optical transfer network OTN), virtual switches (VSI) and virtual networks (VPN, optical virtual private network Ovpn). Virtualized networks can improve network efficiency, such as coordinating bandwidth and topology to meet the needs of specific applications at some point in time. This eliminates the need to aggregate all data center network interconnect facilities to meet peak needs, thereby reducing costs because you no longer need to invest in unnecessary network equipment.

The virtual Datacenter Architecture will provide a "no wall" data center

The

Virtual Data Center architecture, a data center without walls, can help businesses save money and improve operational efficiencies. In companies like Ciena, they provide customers with a "No Walled data center" architecture to create a multiple data center, a hybrid cloud computing environment-as a data center to address many workload requirements and provide seamless workflow mobility. Companies will be able to access cloud computing resources from anywhere, and service providers will be able to provide cloud computing services across different countries, across different programmable network connections and economies of scale. Cloud computing service providers will be able to handle uncertain requirements and failover through more efficient allocation of workloads across multiple data centers. This "backbone" network of cloud computing will be the key to providing cost-effective scalability, security, and on-demand services.