Internet services throughout North America and Europe saw widespread outages and slowdowns on Monday after backbone provider Level 3 Communications suffered a global failure, network providers said.
Time Warner Cable in the US, Research in Motion services for BlackBerry subscribers, and UK ISPs Eclipse Internet, Easynet, and MerulaSupport were all reportedly experiencing problems on Monday. According to multiple accounts, including a variety of Twitter dispatches, at least some of the outages were the result of a bug in Juniper routers that corrupted BGP, or border gateway protocol, tables.
“There are confirmed reports that this event within Level 3 has affected most of North America,” hosting provider Phyber Communications told its customers in a status report. “This outage has affected other networks running Juniper routers with the majority of them seeing their devices core dump and reload.”
A core dump is when a serious error causes a device to crash and the contents in its main memory are written to disk. As a result, the routers must reboot with little or no warning, creating problems for networks that rely on the devices to route email, webpages and other internet traffic.
“Shortly after 9 a.m. ET today, our network experienced temporary service interruptions across North America and Europe apparently due to a router software issue," Level 3 said in a statement emailed to reporters. "It has been reported that a similar issue may have affected other carriers as well. Our technicians worked quickly to address the issue and service is now fully restored.”
In a separate statement, Juniper Executive Vice President Mark Bauhaus said: "This morning, Juniper learned of a Border Gateway Protocol edge router issue that affected a small percentage of customers. A software fix is available, and we've been working with our customers to immediately deploy the fix."
Bauhaus didn't provide technical details, but a purported Juniper advisory leaked online said the bug affected Juniper's MX series of routers.
"The trigger for the MPC crash was determined to be a valid BGP UPDATE received from a registered network service provider, although this one UPDATE was determined to not be solely responsible for the crashes," the advisory stated. "A complex sequence of preconditions is required to trigger this crash. Both IPv4 and IPv6 routing prefix updates can trigger this MPC crash."
The general internet instability was also noted by members of the North American Network Operators Group.
“There are widespread issues across the Internet; certain versions of Juniper firmware have core dumped after seeing a particular BGP 'UPDATE' message,” one member wrote. “It's affected multiple service providers, globally, not just those connected to TATA.” ®
This article was updated to revise the meaning of core dump.
