VeriSign Investing $300 Million Into Internet DNS
Apollo Project aims to provide a thousand times more DNS capacity than today, as VeriSign looks to scale the Internet for the next 10 years.
How much capacity will be needed in the year 2020 by the Internet's key DNS
With Apollo, VeriSign is aiming to dramatically scale up the Internet infrastructure that delivers DNS from its current levels by a factor of a thousand. Doing so will help it manage the estimated 4 quadrillion queries it'll face per day in 2020.
The Apollo effort follows on the heels of the successful completion of VeriSign's $100 million Project Titan that aimed to improve DNS infrastructure tenfold over 2007 levels. Project Apollo thus is designed to move beyond Titan, to help VeriSign meet the Internet infrastructure challenges of the next decade.
"Apollo is about changing and taking to a whole new level the way that we do DNS and the way we do our infrastructure," VeriSign CTO Ken Silva told InternetNews.com. "While we'll still be expanding and hardening the infrastructure, it's also about doing research and development and investment into the future Internet -- whatever that is."
Silva noted that 25 years ago, when the first dot-com address was registered, no one knew how popular the Internet would be by today. As a result, it's difficult to imagine where the Internet will be 25 years from now, though Silva said that it's clear the Internet of tomorrow will be very different.
For one thing, as the global electrical grid moves to an Internet-enabled smart grid, and as increasing numbers of devices get on the Internet, it's no small challenge for the infrastructure to scale.
"The whole reason behind Apollo is that we're at a transformational period in the Internet," Silva said. "For the last 25 years, we've measured the Internet by mouse clicks and e-mails. The next 25 years will be about machines getting onto the network, communicating with each other in ways you don't even know."
Getting there will also require VeriSign to make substantial investments, however -- at least three times the investment it originally made in Project Titan.
"We can't forecast the exact cost, but we do know that it will certainly be more than $300 million over the ten-year period," Silva said.
What $300 million buys
As it stands, VeriSign's current infrastructure can scale to handle 4 trillion DNS queries a day. In 2020, at the completion of Project Apollo, the plan is to be able to handle 4 quadrillion queries, or 1,000 times more.
Expanding capacity isn't just about adding more servers or Internet backbone transport, it's also about improving DNS technologies. VeriSign uses the Atlas DNS server, which it developed.
"We've already started the next generation of Atlas and rebuilding it from the ground up," Silva said.
Silva also noted that VeriSign will rely on the natural evolution of networking technologies to help it make Project Apollo a success. Networking technologies evolve continuously, with one key example being the current evolution of Ethernet from 10 gigabit-per-second rate (GbE) to the new 100 GbE standard, which is coming this year.
Project Apollo will also aim to overcome the challenge of scaling to meet the massive DNS lookup volume that VeriSign anticipates by 2020. In some ways, it's largely a database problem, and VeriSign expects to meet that challenge not with the help of a traditional relational database (RDBMS) "There are networking and compute technologies which have yet to be invented that we will take advantage of by ," Silva said. "Certainly with what we have on hand today, we will apply our best computer science to take performance even further than we've been able to take it so far."
"There are networking and compute technologies which have yet to be invented that we will take advantage of by ," Silva said. "Certainly with what we have on hand today, we will apply our best computer science to take performance even further than we've been able to take it so far."