Is Google Fiber Going to Solve Your Global Connectivity Issues?


The most talked about ISP is finally coming to the Silicon Valley. Google Fiber was approved to install 2,300 miles of their super-powered fiber throughout San Jose, CA. After conquering tech hubs such as Austin, TX, the 1000Mpbs powerhouse has proven it has what it takes to compete with telecom giants such as AT&T and Comcast.

This is a milestone for consumer usage of the Internet, but does it really provide a breakthrough for large corporations or SMBs? Greater bandwidth improves throughput on a local scale, but much more is needed to connect globally and avoid packet loss and latency.

Bandwidth is not the answer to improving your network. In fact, higher consumption might lead to more choke points. And network hardware problems and long distances will add packet loss and latency.

Slow Performance Is Not a Bandwidth Issue
Faster connections, as promised from Google Fiber, will enable the home office to be up and running at full speed locally. The networking term for the local connection is the “last mile”. For the most part, service providers in this connection will get applications delivered with adequate bandwidth and minimum peering issues.

However, there will continue to be latency issues as applications travel across long distances, and across multiple service provider peering relationships. This is no small task, as sending packets around the world using the public Internet requires the cooperation of different service providers and exchanges. Most times, they don’t play well together.

Problems can occur when the data scales get tipped, and one peering service provider gets more than its fair share of data passed to it. Mission-critical SaaS applications such as and Microsoft Office 365 are sometimes hosted in data centers far away from users, and must continually pass through these peering relationships. If these applications happen to travel over a congested network, the performance will degrade.

When applications travel beyond the last mile, which is often the case, they can be sent around the world, and across multiple ISP peers and exchanges. The network term for this is the “middle mile”. It is this middle mile where most network problems reside. The problem stems from a lack of control over long distance networks, where a “free-for-all” ensues with every kind of application competing for the same limited bandwidth. By the time the data or applications arrive, you’ll be thankful they get there safely and without delays.

Improved Application Performance Demands Much More Than a Stable Network
MPLS had been considered a reasonable solution to solving congestion problems, but that level of global connectivity is incredibly expensive and takes months to implement. A network team still needs to prioritize how the bandwidth is regulated.

Furthermore, MPLS does not “improve” application performance by a huge amount as it still suffers from the “speed of light” limitation, a physical upper limit that determines how quickly data can be transferred across an MPLS link. Global businesses are forced to rely on WAN Optimization Controllers (WOCs) to even partially overcome this limitation by enhancing throughput (and therefore, application performance).

Businesses are discovering that MPLS is not worth the huge investment in OpEx and IT resources, let alone the fact that it simply isn’t meeting the network agility and application performance requirements needed today.

Technology Needs to Be Solved from the Middle, Not the Edge
Internet quality and speeds have improved globally over the last decade. Public Internet provides a reliable, agile, and flexible user connection, especially on the first and last mile of the network. The middle mile, over long-distance international links, is where the issue lies. This is where you need the reliability of a global private network, especially when information needs to cross oceans.

Instead of investing in an exclusive end-to-end connection, nearly every global presence could benefit from an MPLS alternative such as a global SD-WAN that empowers users from the middle, thus delivering faster application delivery and performance with minimal packet loss. Not to mention at much greater availability and lower cost.

Many SD-WAN providers can integrate their users into a fast and reliable network at a reduced cost, but only on a regional level.

Aryaka’s Managed SD-WAN solution goes the extra mile, at the middle mile. Instead of your business relying on the public Internet for the middle mile, our global private network is built to deliver MPLS-grade connectivity worldwide. Our global software platform includes WAN optimization and SD-WAN functionality, ensuring a consistent user experience. Global businesses can have a clean and managed network deployed in a matter of days or hours, as opposed to weeks or months with MPLS.

Make no mistake, Google Fiber will be a welcome installation to Silicon Valley, and provide healthy competition for other IPs currently on the market. However, other services such as SD-WAN and global private networks should also be considered if the goal is for optimum worldwide performance. The combination of fast, public bandwidth and a Managed SD-WAN solution would deliver the high performance businesses are looking for at a fraction of the cost.

Want to optimize your network with a Managed SD-WAN solution, give Aryaka a try!

About the author

Mouli Radhakrishnan
Mouli is Vice President of Product Management at Aryaka Networks, and leads the technology process for strategic customer opportunities and business development partnerships.
CTO Insights

2023 Enterprise Network Transformation Report

Download Report >>

Unified SASE Whitepaper

Download Whitepaper >>

Where Do I Start With SASE Evaluations: SD-WAN, SSE, Single-Vendor SASE, or Managed SASE?

A Gartner® Emerging Tech Report

Download Report >>