Johannesburg, 05 Oct 2018
We're occupying a world that's digitally transforming around us. People and things are becoming increasingly connected. Trends such as Industry 4.0, digital transformation and mobile streaming are all producing large quantities of data that require super-fast processing, uninterrupted availability and extreme safeguarding, says Jacques Klopper, Managing Director of Rittal South Africa.
All industries, from healthcare to mining to finance, are using smart applications that demand short latency, the uninterrupted availability of data and system-wide data security. Besides that, there are also industry-specific data storage requirements, such as the need for a high level of physical security for data centres located in harsh manufacturing or mining environments, for example.
The upshot of the move towards connecting everything and everyone is that a massive amount of real-time data is being generated that needs to be gathered, evaluated, stored and made available when needed.
In today's fast-paced digitalised business environment, data has to be processed quickly to enable immediate decisions by cognitive systems using data analytics and machine learning. Once it's been processed, most of the initial data can be disposed of, with only a subset of the volume of data requiring long-term storage.
Where many businesses go wrong, according to Klopper, is in trying to store all of the amassed data in their data centre, instead of filtering it prior to storage. This is where edge computing comes into its own.
He describes edge computing as the decentralised processing of data generated by IOT devices, literally at the edge of the network and close to the application. An edge data centre sits between the cloud and the business's network, and reduces the burden on the cloud or data centre repository by filtering out the data that's relevant to the task at hand, before it's sent on. This means the surging amounts of data produced by IOT devices can be processed and filtered close to the source, reducing the burden on the cloud or enterprise data centre.
There are multiple benefits to processing data at the edge, including a reduction in latency because data doesn't have to be routed via the cloud or physical data centre during processing. It also enables more real-time data analysis, essential for people interacting with IOT devices in sectors such as finance, healthcare and telecommunications.
Implementing an edge data centre also adds an extra layer of data security as it provides more backup on-premises and in the cloud.
Key requirements for an edge data centre, according to Klopper, include the following:
* The system must be modular and scalable so it can accommodate the anticipated explosive growth of data; IDC forecasts a 10-fold rise in worldwide data by 2025.
* Proper cooling of the data centre equipment is key in order for it to run optimally. The cooling system has to be designed with a clear understanding of the current thermal load as well as the projected future thermal load.
* Physical security is paramount. This includes controlling who can access the data centre as well as protecting it against physical threats such as smoke, water and other environmental hazards.
Read more about edge computing here.
Share