If you encountered problems connecting to the Internet on Tuesday [Aug. 12], you weren't alone. Networking experts blame the wide-scale slowdown on outdated routing systems that are reaching their limits as the Internet continues to expand.
People in the IT industry have been referring to Tuesday's sluggish connectivity as 512K Day. Similar to the Y2K worries that caused computing troubles in the transition from the year 1999 to 2000, the 512K problem can be blamed on the arbitrary numbers programmers are often forced to work with to keep systems from becoming too complex or unwieldy.
Just as programmers in the 1990s tried to limit the strain on computers by using just two digits to represent years (i.e., "96" for 1996), the IT industry awhile back settled on 512,000 (512K) routes as an arbitrary limit for the number of routes that Internet routers could handle. Today, the Border Gateway Protocol (BGP) Internet routing table -- a collection of routes that ISPs use to direct traffic through the Web -- has reached that limit, and older routing equipment can't handle the strain.
The result is a slowdown much like the one experienced by many users Tuesday.
It Will Keep Happening
Among those feeling the Internet's pain was eBay, whose users -- at least across northern Europe -- lost access for the 10th time this year, according to a report in The Telegraph. The Web Host Industry Review reported that network providers experiencing outages included AT&T, Comcast, Time Warner and Verizon.
Many Internet users shared their connectivity woes on Twitter using the hashtag #512k.
"Glad to see we are still susceptible to a high impact limit measured in kilobytes," Tweeted IT professional Omer Ayfer. Network engineer Andy Brown commented, "Total routes: 526483 -- Internet problems illustrated! #512k (PS: It's not a day, it'll keep on happening for a while!)."
The problem of "resource exhaustion" at 512K is one that the IT industry has seen coming for a while. Omar Santos, senior incident manager of the product security incident response team at Cisco, warned about the looming issue in a May 12, 2014, Cisco support forum blog post.
"Since the early 1990s, we've watched as the number of entries on the Internet routing table has steadily grown," Santos wrote. "It wasn't that long ago (2008) that the table reached 256k routes, triggering action by network administrators to ensure the continued growth of the Internet. Now that the table has passed 500,000 routes, it's time to start preparing for another significant milestone -- the 512k mark."
A More Fragmented Web Topology
We reached out to David Belson, senior director of industry and data intelligence at Akamai, for his thoughts on "512K Day."
"While many network admins and others in the industry knew about this impending issue, it certainly didn't get the kind of coverage/hype like Y2K did, so knowledge of it may have been insufficiently broad," Belson said in an e-mail. "Ultimately, it's going to require providers and enterprises to inventory their routing hardware and their configurations to determine whether or not they are vulnerable. And they may not do this until the '512K bug' causes problems for their connectivity."
In his May blog post, Santos noted the 512K problem can be addressed through several workarounds. Possible solutions include updating older routers or, barring that, "changing the default configuration for affected devices," he wrote.
Santos added, "There is no related security vulnerability, and it cannot be easily triggered by a remote, untrusted user."
Adrien de Beaupre, a volunteer incident handler at the Internet Storm Center, speculated it was possible there was "malicious" intent behind the routing issues on Tuesday. However, the larger problem remains "increasing scarcity of IPv4 space [as] registrars and ISPs assign smaller and smaller netblocks to customers, leading to a more and more fragmented topology." Internet Protocol version 4, which also is bumping up against a limit on total number of IP addresses, remains the most common network-addressing version used today.
Belson added that a long-term fix would require a meaningful shift to IPv6 over time. "The providers also bear some responsibility here, in terms of announcing fewer, more-aggregated routes, instead of lots of more specific ones," he said. "That will help reduce the size of the table, although this is really more of a bandage and not a complete solution."