Sometimes information sent over the internet yields a blank page or an error code, and though it can be attributed to a user mistake, it is possible that correctly-sent information falls into internet black holes. According to
researchers at the University of Washington, more than 10% of the internet is affected by these strange occurrences, whereby information is lost in transit from one computer to the other.
Whether it’s an email or a Web page request, the information is diverted to the wrong location, and it is lost forever.
Ethan Katz-Bassett, a graduate student at the university, along with his advisor Arvind Krishnamurthy have designed a program to track these black holes with funding from the National Science Foundation. They call the program Hubble, after the space telescope, because of their abilities to recognize black holes in the Internet and the universe, respectively.
They launched the program in September hoping to collect information that would improve the functionality of the Internet, a good plan considering the growing amount of sensitive data that travels across Internet connections.
Though they expected to find black holes, they did not expect to find as many as they did.
"We were astounded when we did an initial four-month study and we saw how many problems there were," Katz-Bassett told Msnbc.com. "It seemed infeasible that this could be happening so often. They’re definitely more common than we thought," he said.
According to the official website, Hubble has identified 904,678 black holes since September 17, 2007.
During a 15 minute run of the system conducted today, April 16, 2008, Hubble detected 2,873 areas of the internet that were unreachable from all locations and 1,030 areas that were only accessible from certain points but not all.
The massive amount of black holes discovered by Hubble shows that the internet is “orders of magnitude less reliable than the telephone network right now,” according to Katz-Basset.
Hubble tracks these problems and makes a real-time map of where they occur.
By charting the areas where the black holes exist, the researchers hope to help internet service providers (ISPs) improve the service to their customers.
"Network administers are definitely interested in it," Katz-Bassett said.
Because most of the lost information happens en route to its destination, they are looking at routing problems as a primary cause of internet black holes.
Routing problems can range from new routers not functioning properly to “multi-homing” techniques that enable a Web site to have several different names/addresses for the same destination.
The information Hubble gathers will be presented at the Usenix Symposium on Networked Systems Design and Implementation next week in San Francisco.
Hopefully their findings will strengthen the internet and make it more reliable, as personal data security is a growing concern among internet users.
By Danny Scuderi
For comments on this article,
email us at firstname.lastname@example.org