9 things I like about Juniper Networks
How many times have you had a bad experience when working on the computer? A page that would not load, errors popping up, bad WIFI quality… You keep on hitting the refresh button hoping that someone will notice that you are having problems because you reloaded that page already a 1.000 times. But no one ever contacted you to help with your problems …
Well, this can change with a Juniper Mist solution. A network environment based on Juniper’s Mist AI will constantly monitor ‘user experience monitoring’ and ‘service level expectations’. Experience monitoring is an ALWAYS ON functionality and is constantly checking that your user experience is satisfactory. The system does this automatically for every user and device and will even raise a helpdesk ticket when it detects performance issues.
Mist AI uses machine learning and AI to make accurate predictions about problems on the network. This is key because network admins do not have enough time to investigate every little problem. Therefore Mist AI will save time and money by efficiently detecting ‘real’ problems.
Marvis, your virtual network assistant
Marvis is a digital network assistant that helps you query and discover problems in the network. It includes:
- A conversational interface
- Marvis Actions
Some people like to barge into the IT office and ask the network administrator ‘what is wrong with the network’ or ‘what is wrong with device x or user y’. Most often, this will be a difficult question to answer unless you are talking to a full-on Network Operation Center. Building such a service, and getting it up and running, requires a critical mass and will take months or even years. On top of that most organizations don’t have the time, nor the staff, to do this.
Juniper, driven by Mist AI, can do all this monitoring from day 1 and will evolve as the network grows. Marvis will be your assistant to whom you can ask questions such as ‘what is wrong with …’. Marvis will use all the data gathered by the platform, the insight brought by AI, and can quickly pinpoint a specific issue.
The ‘Marvis Actions’ are like a ‘todo’ or ‘must look at’ list. This is not a static list, it is constantly being updated and is always up-to-date. Marvis will give you an actionable insight that helps you fix the problem very quickly.
A clean and modern user interface
The Mist AI User Interface is intuitive and gets the job done in a minimal number of clicks. The interface is organized in such a way that you quickly find the information that you are looking for without overloading you. If you want to know more, you can almost always drill down several levels to get more details.
Configuring network services is a refreshing experience as well. Many configuration pages give you a good overview of the parameters you need to look at. To change something, just point and click, and a dialog pops up. It is a modern interface where you will mostly stay on the same page while editing the configuration. This is important because switching pages is often the best recipe for creation of non-consistencies.
Location services and virtual beacons
One of the most remarkable features of the Juniper access points are the built-in location services. It is easy to deploy and manage location services from the same UI. The AP’s feature a unique vBLE 16 antenna array that you will not find with other vendors. BLE technology is used by mobile applications and assets. Since the beacons are included in the access points, there’s no need to place physical beacons, never a need to replace the batteries nor a need to go onsite to install/move/remove them.
Automatic rollback
Losing access to a remote device after making a configuration mistake is one of the biggest fears of an IT administrator. Exactly that functionality is crucial for a cloud management system. All devices are remote devices and you never want to lose connection to the device after sending a bad configuration change.
Mist AI always uses a so-called ‘commit confirmed’ on Junos EX devices. If the device is unable to contact the cloud after the change, it will automatically roll back to the previous working configuration.
API access
Mist AI has an API-first strategy. This means that the API can do everything the UI can and more. In fact, the Juniper API contains features that even the UI does not have yet .
Configuration templates and inheritance
The configuration is structured in 3 levels: Organization, Site, and Device. If you configure settings on the Organization level or Site level, the setting will apply to many devices at once. This is a huge time saver. Whenever needed, a setting can be overridden for a specific site or device.
Another useful feature is the possability to define wired and wireless service settings in configuration templates. These templates can be applied to the entire Organization or specific sites or site groups.
Rapid delivery of new features
With traditional solutions the addition of new features, updates or upgrades and security-patches often result in a painful upgrade process with sometimes risks attached. In practice, this might mean that the software upgrade is postponed by weeks or months and new features are not available in the meantime. Certainly, in case of security-updates, the speed of applying them is key. With a cloud-centric microservices architecture, Mist AI is able to deliver new features as fast as every 2 weeks without having any upgrade issues!
Integration of ChatGPT and LLM into Juniper’s Marvis
In the past few months, the interest in AI has completely exploded with the introduction of ChatGPT and Large Language Models (LLMs). While LLMs represent a huge leap forward in the field of AI, it’s dangerous to assume they replace the need for existing AI-driven virtual network assistants like Juniper’s Marvis. A generative conversational interface makes a tool easy to use and extraordinarily insightful in their replies. However, their answers are based on data captured during a specific snapshot in time, and they can be easily influenced to provide incorrect answers. Therefore, they excel in historical queries rather than real-time issues and makes them currently not suited for real-time insight and troubleshooting.
Juniper has decided to augment the conversational capabilities of their existing Virtual Network Assistant (VNA) and to enable more natural and human-like conversations by integrating OpenAI's ChatGPT API and LLM into its model. This will improve the generation of natural language text, reduce the time spent searching for specific details and provide clear answers. With this integration Marvis can respond to user questions like technical documentation and other publicly available historical knowledge base information. This will help to retrieve information very quickly.
NAC
Network access control (NAC) is a set of policies, applied on a switch, to enforce security so that only trusted users and devices are granted access to network resources. NAC also monitors and controls the activities of the users and devices after network access is granted.
Traditional NAC services are built on on-premises overlay hardware and are often challenging to deploy, operate, and to authenticate devices. They also lack agility and scalability. Therefore, Juniper has recently announced their new cloud-based NAC solution driven by Mist AI. In this way Juniper will not only compete with industry leaders such as Cisco Identity Services and Aruba Clearpass, but also set a new baseline with a very disruptive approach and clear focus on improving user experience.
Juniper's new NAC solution seamlessly integrates with its networking and security services, including wired, wireless, and SD-WAN portfolios. Mist AI empowers the NAC service to gain insights into user experiences and network issues. It streamlines configuration, troubleshooting, and operations through automation.
