- Getting a Move On: Do People with Smartwatches Exercise More?
- Exploring The Frontiers of Cloud Native
- Why Smaller SSDs Are Slower Than Larger SSDs
- The Best Wi-Fi 6E Routers of 2022
- Is a 240Hz Monitor Worth It?
- DisplayPort vs. HDMI: Which Is Better?
- Laptop Fuel Cells Were the Next Big Thing: What Happened?
- Do You Need More Than One Graphics Card for Gaming?
- Our Green Future: Why a Digital-First World needs Zero-Carbon Data Centers
- The Best USB Flash Drives of 2022
Why Autonomous Data Centers Are Inevitable
If you want to predict developments in the IT industry, it might be worth taking a look at long-established industry fields. I’m referring to mechanical engineering — after all, I studied that together with electrical engineering and electronics.
Mechanical engineering initiated the first and second technological revolutions. In this sector, the focus is on increasing efficiency.
Manual production used to be typical in mechanical engineering. This was later transferred to manufacture, where the work steps were divided among specialized teams. After manufacture, assembly lines were established for production. Production steps were further refined, so that only a few activities — sometimes only one — were carried out at each station. These individual steps were later modernized with the help of robotics, creating today’s fully automated production lines.
And now the process in IT has come full circle again. With Industry 4.0, individual production stages are no longer considered. Instead, entire value chains are optimized through digitalization. Autonomy is an essential goal of this where the system makes essential decisions independently.
A good example is the purchase and delivery of an industrial product such as a car. First, the desired model is configured and ordered via a self-service portal. The configuration data is sent to a factory, the car is built, a shipping company is hired, and the vehicle is delivered to and paid for by the customer.
In theory, apart from the customer, no human is required throughout the entire process. However, this means that essential decisions must be made independently. Examples of this are the customer’s creditworthiness, the availability of the necessary components, the available production capacity, and the delivery location and time.
Now let’s compare that with the development of IT infrastructure. Specifically, the essential physical parts of a data center.
First, the individual components of compute, network, storage, and backup were manually configured and operated by the administrator. Later, specialization took place in which the technology was divided into server, network, storage, and backup administration – more or less like manufacturing.
Then, the individual areas were automated, establishing the cloud computing environment we strive for today. This allows services to be ordered via a self-service portal before they are automatically provided and billed. This corresponds to the automated production lines in vehicle construction.
So, What’s the Next Step?
Autonomy, i.e. Cloud 2.0 — or, I would even go as far to say IT 4.0!
- IT 1.0 mainframe
- IT 2.0 client/server
- IT 3.0 cloud
- IT 4.0 Autonomous IT
In fact, many steps are still performed manually in cloud computing, including:
- Patch management
- Resource optimization
- System expansion
- Hardware replacement
- Resource planning
- Further development
And I haven’t even considered the building, power supply, air conditioning/ventilation, and racks – we still have plenty of room for innovation.
What Does Huawei Contribute to the Autonomy of the Data Center?
Now’s the time to get excited – and my IT heart is jumping for joy. ???
Let’s start with the building. Many people do not know that we have set up complete data centers for our customers. Take a look at our data center and the different data center trends here. What you’ll find under trend 5, Full Digitalization and AI-Enablement, sounds exactly like IT 4.0.
And that’s exactly what we have already developed with the use of IoT and AI. The result is an increase in efficiency, demonstrated by shortened availability times, noticeably lower energy costs (electricity and air conditioning), higher stability, and lower staffing requirements, among other things.
My hobby is researching IT infrastructure. So, let’s take a look at some of the important components of data center infrastructure. We’ll start with the network. Everyone has experienced network failures. A component is configured or updated and everything is in place. Even the big hyperscalers regularly make headlines for central network service failures.
To overcome this challenge, we have our AI fabric, which allows the network to optimize itself, detect errors and immediately correct them using AI.
AI-based software even supports planning, installation, operation, and monitoring and thus prevents errors before they happen. To do this, we use a digital twin, which is automatically created from real-time data. The result is increased efficiency, no packet loss, low latency, higher availability, fewer staff, and lower costs. For those interested in learning more, the study Leveraging the Autonomous Driving Datacenter Network Index is worth a read.
We also use similar technologies in the storage area. What distinguishes us from others is that we not only carry out predictive repairs or optimize individual tasks in the storage system, we have a fully automated cycle that generally no longer requires manual intervention. Data is collected in the cloud and used online to train the AI model. The respective model is then transferred to the management systems and storage controllers and executed on dedicated AI processors (inference). Our storage systems learn from each other and become more and more intelligent over time. That’s real autonomy!
We also use this technology in the computer sector, where we are designing more and more processes autonomously. Here we rely on preventive repairs and can predict errors with an accuracy of 93% and up to 30 days in advance. A very interesting solution is how we regulate the energy consumption of the computer systems depending on the load. Not only are the processors throttled — everyone can do that — but the energy-intensive components are put into sleep mode. This not only reduces energy costs but also protects the environment.
These are just a few examples of how we bring autonomy to data centers. Sure, there’s still a lot to be done, but we have already achieved a great deal. Huawei has a decisive advantage here. Not only do we produce cutting-edge technology, we also broadly apply it in Huawei Cloud. At the same time, we are getting ideas for the next innovations from the cloud operation. For our customers, this means that they get tried-and-tested, efficient cutting-edge technologies that they do not have to develop first.
Read more about Huawei’s Intelligent Data Center Solutions.
Article Source: HuaWei
Article Source: HuaWei
Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.