Make the network run autonomously like a driverless car

[51CTO.com original article] Juniper Networks Global CEO Rami Rahim once said: "Many advances in the network field are driving the arrival of 'digital convergence'. For Juniper Networks, we believe that autonomous networks will promote and enable the deployment of 'super services', thereby changing the experience of enterprises and consumers and connecting with users, services and information."

So, what is a self-driving network? How to achieve a self-driving network? With these questions, the reporter recently interviewed Mr. Rami Rahim.

[[195686]]

Rami Rahim told reporters that the so-called autonomous network can be understood as a type of network that can free network operators from complex daily management work, allowing them to use limited resources for more important task optimization. With little or no human intervention, network configuration, network monitoring and maintenance can be performed. Just like a self-driving car, it can free the driver to sleep, read or even work during the journey, or do other things.

Rami said that if you want to achieve an autonomous network, there are some technical building blocks that need to be in place:

The first is network telemetry technology. The network needs to have a high level of capability and be able to provide telemetry technology to a large extent. A large amount of data and information needs to be collected in the network, such as network operation status information and network performance information. All network components need to be able to provide such information. One of Juniper Networks' innovations is to improve the efficiency of network telemetry technology.

The second is machine learning technology. Through the machine learning platform, we can transmit data, absorb, learn and understand the data provided by the current network, so as to understand the current status of the network, and then allocate network resources based on this information, optimize network resources and so on.

The third is to enable the network to have self-debugging capabilities. Through self-debugging capabilities, configuration can be made in the entire network and deployed securely. In this regard, Juniper Networks also provides solid products, such as the SDN controller, which is responsible for providing corresponding functions for the cloud network. There are also products responsible for wireless services.

Fourth, people’s awareness. People’s acceptance of this autonomous network, whether their skills are ready, whether they trust this technology, and whether they believe it can really improve efficiency. Just like when we first come into contact with self-driving cars, we may have doubts. We need to slowly accept it and believe that it can drive itself. This process takes some time.

The autonomous network can achieve high-end intent basic service description, end-to-end device autonomous service management, optimized telemetry basic service deployment, and real-time service demonstration. The automated management of the autonomous network mainly includes five stages: manual control, virtualization, prediction, suggestion, and autonomous operation.

In addition, in the interview, Rami Rahim said: "Juniper Networks is a company that focuses purely on network services and technology enhancements and is an innovator in the network industry. We have been constantly looking for technologies and solutions to solve some critical network problems."

Later, he told reporters that Juniper Networks' innovation now has two dimensions: First, further improving the dimension and performance of network expansion. Since 2016, Juniper Networks has launched a group of new products around the data center switch market and core router products, hoping to extend the length of the entire IP network through this product. Second, automation. Through the promotion of automation to meet greater needs, it also promotes Juniper Networks' services to be more agile and allow services to reach end users faster. In addition, we also improve cost-effectiveness and manage large and complex equipment and networks.

According to the reporter, in order to achieve network automation, Juniper Networks recently launched a machine learning platform that uses artificial intelligence algorithms to achieve intelligent management and operation of network and cloud resources, while ensuring that cloud services never go down.

In an interview with ***, Rami Rahim revealed to reporters that Juniper Networks' performance in the first quarter of 2016 and 2017 was very impressive, especially the performance of data center switches, and the growth of cloud service products reached more than 20%, and many of its customers were from large global cloud service providers.

[51CTO original article, please indicate the original author and source as 51CTO.com when reprinting on partner sites]