Edge Solutions: To the Edge and Beyond
The demand for big data and instantaneous video, IoT growth and the explosion of social media are pushing IT management teams to consider edge solutions, where computing occurs closer to the actual source of the data itself. Innovative, disruptive technologies, such as driverless automobiles, industrial IoT (IIoT) and artificial intelligence (AI) systems will only accelerate the need for edge solutions in data centers.
Other new applications drive this need for edge solutions as well:
- Auto manufacturing plants that utilize robots and AI
- Automated toll booths, where motion sensors, cameras and software work together to detect vehicles and license plate data, and gather payment information
- Biopharma, which makes use of efficient just-in-time manufacturing and highly technical processes
- Healthcare applications, including wearables and AI engines
- Petro-chemical applications that use sensors and IIoT
- Smart electrical grids
- Water/wastewater treatment applications that make use of sensors to capture data
All these applications have one thing in common: They have a need to compute and store results and data locally at an edge facility. Then, as required, they can transfer data/results back to a central data center or to other facilities.
The Definition of an Edge Ecosystem
Edge solutions provide data services (whether it involves content, compute or storage solutions) at the last mile. By creating data at a source, using data at a source, sending data to a local data center (the edge facility) and sending or receiving data from a larger, central data center if needed, you can reduce the amount of data that must flow back and forth to improve speed.
The Nuts and Bolts of Deploying an Edge Solution
There are several key stakeholders involved with deploying an edge solution, from cable and telco companies to the C-suite, facilities engineers and end-users.
Stakeholder teams managing edge data center programs have the same desires as an enterprise/corporate IT department; however, due to speed of deployment, size, scale and flexibility, the edge team needs to be agile and innovative to take full advantage of the data center technologies designed to provide an efficient, functional edge solution.
Developing an edge solution is complicated, and requires participation from all stakeholders. Before kicking off an edge data center program, however, there are four simple questions the stakeholders need to address as a team:
- What type of IT service/load will be deployed at the geographic location?
- Where are these geographic locations?
- What types of data and critical equipment redundancy are required?
- How many edge facilities are needed?
Answering these questions will help you develop the best roadmap to follow and pinpoint project objectives prior to digging into technical design/construction/deployment issues that could interfere with your primary goals.
How Traditional and Edge Data Centers Compare
Today’s edge data center operations contain all of the elements that serve a traditional data center:
- Scaled-down critical infrastructure
- Mechanical/electrical/fire protection systems
- Redundancy requirements (not Tier 4)
- Networking solutions
- Connectivity options and redundancy
- Shell building
- White space/mechanical space/electrical space
Due to their smaller data center footprints, speed to market, diverse compute requirements and low latency/high-speed transmission requirements, edge facilities require unique design, construction and operation compared to a traditional enterprise or multi-tenant data center:
- Just-in-time deployment
- Built-in security systems
- Software-defined data center redundancy
- Considerations for power usage effectiveness (PUE) vs. uptime
- Operating a dark/lights-out facility with minimal human access
- Remote-monitoring capabilities
- Modular electrical power solutions
- Scalable cooling solutions
- Predictive DCIM systems
- Innovative/efficient network storage systems
Where is Edge Headed?
The demand for edge solutions will certainly continue to grow as IoT, social media and big data analytics expand, and consumers develop their insatiable desire for faster media-download times and more online content.
Edge data centers as we know them right now may continue to change as well. For example, you may begin to see these cool new technologies deployed with edge solutions:
- Micro data centers
- Rack-scale design
- Dynamic cooling optimization
- Rack-centered solutions
- Software-defined networking and critical infrastructure
Belden’s data center experts can help you deploy customized edge solutions that satisfy the demands of today and the future. As edge solutions change the way data centers are designed, maintained and managed, we’ll help you manage the impacts.
关于作者
For more than 15 years, Daniel Bodenski has worked as a design professional with world-class data center owners/operators. Daniel held positions with EEC/NTT Facilities, CRB Consulting Engineers and Syska Hennessy Group. His goal was to help data center operators build, maintain and operate a highly efficient, reliable and scalable facility.