数据中心

数字化转型正在“边缘”发生

Shad Sechrist

无论我们是在工作还是出差,甚至是在医院,我们都在使用移动设备消费数据,而且这正以爆炸性的速度增长。

 

To prove this point: Smartphones were used for 61% of visits to U.S. websites last year, according to digital consultant Perficient. Business Insider predicts that 41 billion IoT devices will be connected to a network by 2027. In the next two years, 70% of vehicles will be connected to the internet, says Gartner. As wireless data usage increases, demand for data storage and network throughput is on the rise as well.

 

This surge in mobility—and the amount of traffic being generated by applications like IoT and artificial intelligence—is moving us toward 5G wireless technology, which means that bidirectional (send and receive) traffic is a significant consideration.

 

Relying on one central location to send and receive this data wastes valuable time and money. It can also create congestion and long transmission times that interrupt users and lead to damaging downtime or failure.

 

The solution? More devices will need to send data to the “edge” and the cloud—and we’ll need to rely on more than one type of infrastructure (or a “hybrid” approach) to make this happen.

 

The concept of edge computing involves repackaging traditional, on-premises computing and moving latency-sensitive applications like multimedia streaming and telehealth closer to people for very fast response times. Edge computing is often needed to process data beyond the public cloud or service-provider cloud, pushing the “compute” function out to edge devices. A few examples can be found here.

 

Amid the emergence of edge computing is a sector we call “hybrid edge computing,” which is similar to other “hybrid” terms you may have heard: hybrid cloud, hybrid IT, hybrid computing or hybrid edge data centers, for example. But its location—at the edge of the network—makes hybrid edge computing a little different than other hybrid solutions.

 

Hybrid edge data centers (or hybrid edge computing) support a mixed approach to computing, storage, security and carrier services for efficient and cost-effective data management, storage and analysis. They deliver latency-sensitive data close to users and integrate with core or centralized applications in public clouds or corporate data centers. Instead of supporting only outward-bound data, they also support bidirectional data we mentioned earlier.

 

To better understand the concept of hybrid edge computing, it helps to think about a wheel’s hub and spokes. Although they’re located at different points around the wheel, each spoke connects to the same central location. The place where the spoke hits the outer part of the wheel can be considered the “edge” in hybrid edge computing. A hybrid solution uses third-party services to connect all the components of the edge to all the other components of the wheel. This creates more of a mesh network instead of simple hub-and-spoke network.

 

The Rise of SD-WAN in Hybrid Edge Data Centers

 

As with any new approach to infrastructure, there are some unique challenges associated with supporting hybrid edge data centers.

 

In traditional data center environments, there are people nearby to monitor performance and troubleshoot issues. Hybrid edge data centers are usually unmanned, which means they could be fully automated, remotely managed and operate without staff. This makes remote, plug-and-play edge architecture solutions key, along with automation.

 

For this reason, software-defined infrastructure (SDI) is becoming a  necessity. It supports the operation and control of IT infrastructure through the use of software—and without any human involvement. In particular, SD-WAN (a software-defined wide-area network) has proven to be a valuable tool in delivering certain types of connectivity.

 

SD-WAN and hybrid edge computing can work together to share infrastructure and route traffic to appropriate edge resources.

 

Through SD-WAN, application traffic can be directed to onsite or edge resources with available service, which can be particularly helpful for organizations that have multiple locations spread across a metro area (such as banks, big-box stores or gas stations, for example).

 

SD-WAN can run on edge computing hardware … or act as a platform to support edge computing. It also helps automate moves, adds and changes by decreasing the amount of time and number of resources required while automating network programming.

 

What Hybrid Edge Computing Means for Your Cabling

 

Because of these shifts, your approach to cabling, connectivity and management will also change (of course).

 

The success of a hybrid edge data center comes down to interconnection points (meet-me rooms). These managed spaces allow cable companies, internet service providers (ISPs), telecommunications carriers and other carriers to cross-connect. They support physical connections (mostly fiber) of different companies and ISPs in the same building, allowing them to exchange data before distributing services to the internal network. They also support inter- and intra-campus cross-connects.

 

Meet-me rooms are critical to support uninterrupted internet exchange and smooth data transmission. They connect companies and carriers in the same data center space while removing round-trip traffic and ensuring that data doesn’t have to leave the facility (lowering costs and boosting security as well).

 

In hybrid edge data centers, meet-me rooms handle connections to cloud providers, to the core network and even other to hybrid edge data centers, as well as to disaster recovery or offsite storage locations.

Speak to a Belden Expert by visiting our Data Center Market Page!