When the internet becomes the new LAN
- By Pejman Roshan
For decades, enterprises have invested heavily in their networks to ensure reliable and fast access to applications and data. Millions of dollars have been poured into building robust and highly manageable local- and wide-area networks and data center infrastructure to ensure that users can access data efficiently. Today, though, agencies are funneling their IT budgets into software-as-a-service deployments like never before. In fact, Gartner predicts that by 2021, 28 percent of all IT spending will be on cloud-based infrastructure, applications and business services.
By moving applications to the cloud, agencies are throwing responsibility for performance, visibility and control at a variety of SaaS vendors, internet providers and the internet as a whole.
But are those entities prepared to catch this responsibility and run with it? SaaS providers, of course, have a vested interest in delivering optimum application performance for their customers. They use content delivery networks to accelerate transmission of static website content, and they optimize their code to make web apps as efficient as possible. They share their applications across multiple points of presence, moving applications and PoPs closer to users.
This isn’t enough, though. As we are witnessing, none of this work addresses the core performance issues presented by the internet. For the same reason agencies have always updated and enhanced their LANs and WANs, they should be just as concerned today about internet performance. If the agencies are rightly concerned about internet performance, the SaaS vendors should be, too. They are, after all, responsible for delivering a turnkey service.
It's critical that SaaS vendors deploy the latest advances in internet network overlay technology to best guarantee the delivery and performance of cloud-based apps so critical to agencies' missions.
Internal IT teams will guarantee performance within the four walls of the enterprise, but the unfortunate reality is that SaaS vendors will bear the responsibility for internet performance as it relates to enterprise applications. When performance issues start impacting user productivity, enterprises will look for a throat to choke.
The internet is designed for resilience, not performance. The routing protocol of the internet, Border Gateway Protocol, does not factor in link or autonomous system performance. Subtle yet impactful design flaws in BGP, such as slow convergence time, allow for routing loops that lead to packet loss. In addition, least-cost routing and low-cost interconnects provide no quality-of-service guarantees. Worse yet, these issues all occur in the middle mile -- inside the networks of ISPs and cloud providers -- and away from last-mile connections and links. This means neither SaaS vendors nor enterprises have a toll-free number to call and complain when SaaS applications and websites lag.
And so responsibility lands squarely in the lap of the SaaS provider and, more specifically, the DevOps or site reliability engineering team of that SaaS provider. These teams must be network-literate because the best way to address today's internet performance challenges requires looking to networking-oriented solutions.
Cloud platform vendors offer direct connection solutions, and transit providers will claim bold performance numbers. But today’s continued SaaS performance issues show it’s not that easy. Agencies need employees with solid networking skills to dig into a vendor's claims and discern what is real. Different SaaS applications have different networking footprints and varying degrees of internet performance requirements. One size does not fit all.
Internet network overlay solutions are bringing performance, reliability, visibility and control to the internet backbone connection between the SaaS provider and the enterprise. There are both private network- and public cloud-based alternatives. Some advanced solutions leverage machine learning to track real-time performance across the internet backbone and predictively and dynamically route traffic from the SaaS provider to the enterprise user through the highest-performing path, no matter where in the world the user is.
These overlay solutions are typically delivered as a turnkey service, allowing oversubscribed DevOps teams to concentrate on delivering the best application performance and reliability while the overlay service delivers the best network performance and reliability.
Pejman Roshan is vice president of products at Teridion.