NEW YORK (CNNMoney) -- The seemingly indestructible Internet relies on a few backbone systems to keep traffic flowing smoothly. Sometimes, one of those systems blips -- and millions of devices get abruptly kicked offline.
That's what happened Monday morning, when a software glitch in the Internet's wonky sounding "Border Gateway Protocol" created a ripple effect that crashed data networks around the world.
The outage appears to have originated early Monday with Juniper Networks (JNPR), a company that makes router hardware for large networks. A set of updates to a core Internet routing protocol triggered a software glitch in some of Juniper's routers. When those routers crashed, key Internet pathways went down with them.
Time Warner Cable (TWC, Fortune 500), one of the largest Internet providers in the United States, suffered a short but widespread blackout. "We appear to be recovering from a large but brief internet outage affecting most of our service areas," the company wrote on its customer-service Twitter account.
But Time Warner's problem most likely originated even further upstream: Level 3 (LVLT), a so-called "tier 1" network and one of the key Internet gateways in the U.S., got caught in the crash.
"Shortly after 9 a.m. ET today, our network experienced temporary service interruptions across North America and Europe apparently due to a router manufacturer software issue," Level 3 said in a statement released 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."
While Level 3 didn't name names, Juniper confirmed that its systems ran into trouble Monday morning: "This morning, Juniper learned of a Border Gateway Protocol edge router issue that affected a small percentage of customers," the company said in a statement to CNNMoney. "A software fix is available, and we've been working with our customers to immediately deploy the fix."
Internet monitoring service AlertSite said its systems picked up "a measurable event during the 9-10 a.m. EST hour."
That "event" took down systems big and small. Many BlackBerry users -- already skittish from last month's widespread outage -- found their devices temporarily knocked offline. BlackBerry maker Research in Motion (RIMM) was quick to point out that its systems weren't to blame, citing "a global Internet issue."
Tweet shortening service Bit.ly was another victim, temporarily rendering all of its links unusable.
By late morning, most sites were back to normal, shaking off the sluggishness that typically follows a major routing glitch. Like a human fending off a cold, the Internet occasionally succumbs to bugs, but it also tends to bounce right back. 
