Posts tagged ‘VeriSign’
Serving Over 70 Billion Daily – The Growth of Google Public DNS
Google made some big news in the world of DNS recently. The company announced on their blog that they currently respond to over 70B DNS requests per day, making Google the largest provider of public DNS services. (DNS translates domain names to IP addresses, and returns the correct online destination to users).
Back in 2009 I wrote about the launch of Google Public DNS, and wondered what effect it would have on the market. One result is that consumers now have a well known technology company offering them a better DNS experience than they get from their ISPs, which historically have never had any reason to invest in better DNS performance. Google claims it does not store personally identifiable information from DNS traffic, and does not combine it with any other Google service.
Other industry players will need to respond to the growth of Google Public DNS. OpenDNS is a startup company that in many ways pioneered free recursive DNS service in 2005. (Quick DNS primer — recursive DNS returns answers to users, authoritative DNS provides those answers to be returned).
Forbes just published a Q&A in which OpenDNS founder David Ulevitch attempts to spread some fear, uncertainty and doubt (FUD) about Google becoming too central in how users connect online. Over the years I’ve been impressed with how quickly Ulevitch responds to any market development that could affect OpenDNS. In particular, he started using social media channels to support his corporate messaging well in advance of other companies in the space.
This well-timed media interview is another example. With coverage of Google overriding privacy settings in Safari browsers fresh in the news, it’s smart messaging to raise doubts about the company becoming dominant for DNS.
At CircleID, there’s a fascinating discussion on how Google’s growth affects VeriSign, operator of the dominant .com and .net domains. VeriSign operates .com and .net under an exclusive contract with ICANN, the administrator of certain Internet infrastructure functions. The piece is written by George Kirikos, a long-time critic of ICANN who uses the 70B requests a day number to criticize VeriSign’s exclusive contract.
George points out that one metric VeriSign has used for years to justify its ability to effectively support .com and .net is DNS requests handled. Based on the latest public statement, that number is currently a daily average of 59B, less than Google’s 70B. As the comment thread of the piece makes clear (worth the time to review), the comparison isn’t really apples to apples since Google is recursive and VeriSign is authoritative. Even more critically, VeriSign has a much heavier burden to provide a stronger level of service than does Google.
Most people don’t understand how DNS works, never mind understand the difference between VeriSign’s number and Google. Having divested itself of its telecom business (to TNS in 2007) and security (Symantec in 2010), all of VeriSign’s eggs are in the DNS/.com/.net basket. (The company also has a small managed security business called iDefense). Rather than trying to explain the differences and draw further attention to Google’s growth, I’d expect VeriSign to drop the DNS request number from its external messaging.
As I wrote back in 2009, Google was sure to create big waves by entering the DNS space. It sure makes the space more interesting. And as Ulevitch of OpenDNS admitted in his interview, the growth of Google Public DNS raises the awareness of the importance of DNS. That’s a very good thing in and of itself.
Talking Domain Names with Antony Van Couvering
Today I’m pleased to publish the first of a two part interview with Antony Van Couvering, of Mind + Machines. As you will read, Antony has been a force in the domain name business since before there way such a thing. The recent approval of new top level domains (TLDs) by ICANN promises to fundamentally change domaining, and Antony is right in the thick of that transformation as well.
So there is no better time to get a true insider’s take on the gTLD process. In the early 2000s Antony and I both worked at VeriSign, for different business units. The second part of this interview will be published on Tuesday, July 12.
Q: Can you give a background of your long involvement in the domain name registration business?
I began in 1997 with an email to Jon Postel at IANA to find out what these ccTLD things were. Jon and I began talking and it wasn’t long afterwards that I joined up with an English partner, Ivan Pope, to form NetNames. I ran the U.S. branch, NetNames USA. I quickly began defining what is now called the corporate market among domain registrars, the space now dominated by Mark Monitor and a few others, by contacting and working with many of America’s biggest companies and law firms. In those days people were very wary of the idea of domain names, especially ccTLDs, and were struggling to understand what it meant for their brands. It was pretty much the wild west, and I was able to help by bringing the perspective of TLD managers to companies in a way that made sense to them. Eventually Ivan and I sold the company to NetBenefit, now called Group NBT.
During that time I got quite involved with Internet governance issues, working with Jon Postel to create a proposal for revamping the .US domain and then spending a lot time helping to set up ICANN. As part of that work, I helped form the International Association of Top Level Domains (IATLD), which pushed back very forcefully against ICANN’s attempts to rewrite the ground rules for ccTLDs, and in particular to effectively deprecate RFC 1591. Then as now, the major challenge was dealing with governments, whose approach of “sovereignty” and “rights” is very different from the consensus-based, co-operative ethos that characterized the Internet then, and still does to a large extent.
I then started NameEngine, which addressed the same market and clients as NetNames, and did very well until I sold NameEngine to VeriSign. As part of the deal I worked for them for a couple of years. It was a pretty strange place to me back then; Mark McLaughlin and his team have done a great job to make VeriSign the company it is today.
Having launched several ccTLDs during my time at NetNames and having helped several others, I’ve always been very interested in new gTLDs. When ICANN first approved the development of the program a few years ago, I decided to get involved, and began blogging and consulting under the umbrella of Names@Work.
Then in January 2009 I started Minds + Machines with Elaine Pruis. We got an exclusive license from CoCCA to use their platform for gTLDs, which we rechristened Espresso, and with Garth Miller of CoCCA began developing the platform to meet the evolving requirement of ICANN for new gTLD registries — as everyone else in the industry had to do as well. Soon afterwards, our investor company Top Level Domain Holdings, Ltd., which is a public company listed on the AIM market of the London Stock Exchange, acquired 100% of Minds + Machines, and I’m now the CEO of both Minds + Machines and Top Level Domain Holdings. We’ve raised quite a bit of money and we’re looking forward to participating in the opening of the new gTLD space.
Q: ICANN just approved the introduction of new TLDs, and applications are expected by January 2012. How would you characterize the opportunity this represents, for operators and users?
The opportunity is truly historic. ICANN was set up to make new gTLDs a reality, and it’s taken 12 years of shouting to get it done, after a couple of false starts. While subsequent rounds are sure to happen someday, it’s anyone’s guess as to when: this round took six years to accomplish. For applicants, there’s suddenly a rush to get things done: applications to ICANN have to be submitted during the window that lasts from January 12 to April 12, 2012. That means that applicants have just a few months to put together a pretty complicated and serious application, and to round up the money they need. For users, it’s a great time to get a short, memorable domain name that matches their brand or their name or represents the business they’re in. Right now under .com it’s tremendously expensive, since good names have to be purchased in the secondary market.
Overall, the major benefit from new gTLDs will be to re-introduce semantic meaning into the characters that follow the dot in a domain name. While most country-code top-level domain names have a meaning, either in a geographic sense or as a repurposed gTLD (e.g., .tv), the most popular gTLDs are essentially meaningless at this point: anyone can register anything in .com, .net, and .org, and they do — those three characters after the dot no longer mean “commercial,” or “network,” or “organization.”
One commonly voiced criticism of new top-level domains goes something like, “No-one needs them, everyone just uses search anyway.” Even if you ignore the data that about 15% of all traffic is from typing in a URL or using a bookmark, new TLDs will actually work very well with search. I just wrote about this in a recent blog post. This is just another way that new TLDs will benefit users.
Q: Recently your company announced it was working together with Neustar for certain TLD opportunities. How will you be working together?
We have a two-way exclusive agreement with Neustar for geographical names, with some exceptions for projects we undertook before joining announcing our partnership. The deal is that we agree to use their registry platform for geographical TLDs, and they agree to refer all relevant clients to us as the front-end partner. I think it’s a great deal for both of us: we’ve been busy over the last two years making a lot of contacts with cities, while Neustar has a back-end that is very solid, and a recent success doing the back-end for .CO.
Does this mean we don’t like Espresso anymore? No, it doesn’t — we love Espresso. But as everyone who has been following the progress of the new gTLD program knows, governments are extremely risk averse, and governments are the decision-makers when it comes to geographical gTLDs. To a government, Neustar’s profile as a big public company on the New York Stock Exchange, with a $2B market cap, makes a lot of the queasy what-ifs go away. We are interested in participating to the fullest extent in the new gTLD space, and we’ll do what it takes to make that happen. In this case, we thought the partnership with Neustar made a lot of sense, and we still do.
What the Heck is ICANN Doing with Domain Names?
Last week stories broke about a significant change in the way Internet addressing will be managed. The Internet Corporation for Assigned Names and Numbers (ICANN) has opened up the process of assigning new top level domains (TLD’s), such as .com and .net. Potentially any string of letters could be a TLD — maybe we’ll see www.products.walmart? However there appears to be lots of procedures to iron out before any prospective new TLDs hit the market.
Here’s some coverage:
Ars Technica:
http://arstechnica.com/news.ars/post/20080626-confusion-icann-opens-up-pandoras-box-of-new-tlds.html
NY Times:
Here’s a very detailed read from John Levine at CircleID:
http://www.circleid.com/posts/86299_icann_new_top_level_domains/
Clearly nothing is going to happen right away, and the devil(s) will be in the details. Plus there are already at last count 162 million registered domain names, so a lot will be required to produce a new TLD that actually makes a difference. Here’s my early list:
- User adoption — the old chicken and the egg. An organization with a lot of money and influence will have to invest both to change user perceptions of Internet addressing. It’s a steep hill to climb. Combined, .com and .net total around 85 million domains (with .com about 74M of that) and will be the TLD leaders for a long time to come. (Granted both China’s .cn and Germany’s .de are now bigger than .net)
- You’ll need a solid sales channel. The successful TLD operator will need good connections to the existing base of registars, and a compelling value proposition for registrars to sell the new extension over other, more established ones.
- Global resolution is a big responsibility. The new extension will have to always work, anywhere on the globe. That means servers and data centers in multiple locations, load balancing, maintenance, etc. etc.
- Domain names in foreign languages sound very interesting. But as Levine notes these could get hopelessly bogged down in litigation, and many countries particularly in Asia have been forging ahead with their own fully native language solutions while ICANN has dithered over this question for years.
No doubt this will be a very interesting story to follow. If I had to bet my lunch money today, I’d say that the established players in the registry market who already know how to operate and support TLDs will be the main beneficiaries of ICANN’s decision. They’ve made the technology investments and mastered ICANN esoterica. Any newcomer will have to partner with one of them if a new TLD hopes to attract wide spread adoption.
Five Years Later, a SiteFinder Patent?
Saw an interesting news item that broke Monday courtesy of DomainNameNews and SlashDot that hasn’t been broadly covered yet. Apparently VeriSign has been awarded a patent for the resolution of mis-typed domain names. This was at the heart of the controversy back in 2003 around their SiteFinder Service. Amidst a storm of criticism ICANN insisted VeriSign shut down the service, and the company eventually agreed.
Patent: http://www.domainnamenews.com/registries/verisign-receives-sitefinder-patent/1559
SiteFinder: http://www.itworld.com/Man/2681/031006sitefinder/
Personally I believe if VeriSign had been less secretive about its plans and had briefed important Internet constituencies beforehand about this change to how the Internet operated, there would have been less criticism. VeriSign and ICANN eventually settled their differences re SiteFinder, and as part of the re-awarding of the .com franchise VeriSign promised never to bring a SiteFinder-like service back. But, this patent is interesting for what it means in the present, not the past.
Many companies currently resolve incorrect domain names to pages that contain advertising. They don’t do it the way VeriSign did it, by changing the way the root server operates for .com addresses. But the result is the same, and would seem to be covered by the patent. Companies like Earthlink, Verizon and OpenDNS bring their customers to advertising pages every time they “fat-finger” a domain name. Just think about how often we all do this, and you can imagine the amount of Internet traffic potentially involved. Some of these companies weren’t re-directing this traffic safely — there were security problems with the ad partner Earthlink chose to deal with:
http://www.portfolio.com/news-markets/top-5/2008/04/21/Hackers-Can-Exploit-Error-Page-Ads
So, the government seems to have handed VeriSign a new revenue stream. Letters may be going out shortly demanding a licensing fee — or maybe they’ve already been sent.
UPDATE — AP wrote story: http://biz.yahoo.com/ap/080514/techbit_internet_typos.html?.v=2






Recent Comments and Pingbacks