Will cloud computing kill off VPN & MPLS networking for Business?

Comments are off
Like
Darren Gore
Will cloud computing kill off VPN & MPLS networking for Business?

Virtual Private Networking (VPN) and Multiprotocol Label Switching networking (MPLS) were both originally designed to network together servers, sites, people and devices across states, a country or around the globe.  

The VPN protocol, first implemented in 1996, was developed by Gurdeep Singh-Pall, a Microsoft employee who is now a corporate Vice President. He invented PPTP (Point-to-Point Tunnelling Protocol), as a method for implementing a private network over a public connection. Singh-Pall developed this to allow users to establish a secure connection to work resources from home using a public connection. This was a milestone event that truly set the stage for the evolution of the VPN and in many ways enabled ‘work from anywhere’ to become a reality. 

 

What is MPLS? 

 

According to Infocellar, MPLS as we know it today was established in 1997 by the Internet Engineering Task Force. It was developed as an alternative to earlier multilayer switching protocols and is primarily a method of securing connections directly between whole sites or geographically dispersed offices.  MPLS is a specific type of protocol that enables the transmission and shaping of network traffic and has been a staple of enterprise connectivity since it was developed. 

 

The purpose of both of these networking methods was to allow for secured transmission of data, and in many cases, to apply a quality of service (QoS) so that specific traffic, such as voice and video, could be prioritised. 

 

The Server Era 

 

When Microsoft released Windows NT in the early ‘90s, the methodology for businesses computing was to centralise their business software such as Accounting, CRM, File Management, Email etc. on dedicated central servers. Roaming business users would then access these systems externally using VPN technology or MPLS networking to securely connect other sites back to these central servers. 

 

  

Colocation, Cloud & Hosted applications 

 

Around the year 2000, we saw the first examples of cloud hosting which started as Colocation (CoLo). This is where a business replicated what they had been doing in their own server rooms, but would instead locate their own equipment in a dedicated facility that was shared with others.  This reduced costs by sharing it across a number of companies and enabled higher levels of service, redundancy and security that most companies could not afford themselves. Regardless of the actual equipment location, this still required a secure networking method, either VPN or MPLS, to access those servers. 

  

Introducing IaaS 

 

In 2006 the first iteration of Infrastructure as a Service (IaaS) was launched. Extending on the CoLo concept of shared resources, the IaaS model allowed a business to lease server equipment and resources, thus releasing them from capital acquisitions and the need to manage the underlying hardware. With developments in virtualisation from Microsoft and VMWare, the IaaS model has become a method of abstracting business logical servers away from the constraints of specific hardware. The connections required to this infrastructure however, remained the same – the customer would have servers (now virtualised) which would host the central applications that users would use a VPN or MPLS networks to access that central infrastructure. 

 

The Creation of SaaS 

 

By 2006, Microsoft launched Hosted Exchange Services which would later become part of Office 365. This particular date is important as it was a milestone in the business software world that would trigger a significant change in thinking and set a new direction for Microsoft and business software systems. This new service or Software as a Service (SaaS) as we call it today, is now the current mindset for cloud computing.   

 

Via ZDNet

 

SaaS changes the software distribution model from developing, packaging and distributing/running in many locations to one where the software vendor can develop and run a central optimised instance and then makes that service available to customers. The end user then no longer needs to manage the installations, updates or servers on which that software runs.  

This approach is beneficial for all parties involved. The software developer has more control over the experience, they manage ongoing updates and the hardware platform can be optimal for the application. This allows a consistent customer service experience end-to-end. Additionally, the removal of marketing, packaging, distribution and distributors allows removal of unnecessary costs. 

The variable in this from previous application delivery is the network connection. In the SaaS model, services are delivered directly to the end user over standard Internet connections. They do not require a VPN or MPLS, using protocols like Secure Sockets Layer (SSL) to authenticate and create a secure connection from the user’s device directly to the cloud platform.  

 

What does this then mean for the future of VPN or MPLS?    

 

Since businesses are likely to require multiple connections to different SaaS applications, protocols that connect a user to a single point are less useful – we’ll need secure ways of connecting the user to multiple applications in different locations simultaneously. Therefore, it’s likely we will see less and less reliance on VPN or MPLS networking in the future. What will be important is our connectivity to the internet, the bandwidth and data allowances.  

As Australians, NBN evolves and ushers in higher levels of connectivity and performance, we should see SaaS platforms as more viable for a greater number of businesses than ever. For those businesses who previously relied on expensive MPLS networks spanning states and countries, the ability to use more generic lower cost direct internet connections will change the way they work and help reduce costs.