Brief, global internet outages blamed on software bug

Fastly glitch will continue conversation around security and reliability: Cybersecurity expert Shopify, Reddit and Amazon are just a few of the names that were affected by an outage this morning that introduced many of us to a company called Fastly, which helps shuttle data across the internet. In a tweet shortly after 7am, Fastly pinned the blame on a "service configuration" and said its network was ramping back up. Alex Hamerstone, GRC practice lead at TrustedSec, joins BNN Bloomberg to weigh in on this. He says that the Fastly glitch highlights how dependent so many different sites are on a single service.

A software bug at a major network provider briefly knocked dozens of financial institutions, airlines and other companies across the globe offline during peak business hours in Asia.

Akamai, which runs one of the internet's main content-delivery systems, said the outage Thursday was not caused by a cyberattack, but rather a software bug on a service that protects customers against denial-of-service attacks.

Many of the 500 affected Akamai customers had their traffic rerouted in minutes but it took more than four hours to fully restore the system, the Massachusetts company said. Akamai operates mirrors of customer websites in 135 countries -- known as edge servers -- designed to speed access to them.

The Hong Kong Stock Exchange and the four largest U.S. airlines were among those impacted. Akamai does not name its customers but says they include more than 300 of the world's banks, more than 30 airlines, more than 200 national government agencies and 825 retailers.

Many of the outages were reported by people in Australia trying to do banking, book flights and access postal services at mid-afternoon. Many services were back up and running after an hour or so.

Banking services were severely disrupted, with Westpac, the Commonwealth, ANZ and St George all down, along with the website of the Reserve Bank of Australia, the country's central bank. The Reserve Bank cancelled a bond-buying operation due to technical difficulties facing several banks that were to participate.

The airline Virgin Australia was also affected and cited the Akamai content delivery system. It said flights largely operated as scheduled after it restored access to its website and guest contact center.

Outages briefly spiked on American, Delta, United and Southwest airlines. Because the disruptions happened late at night in the U.S. when few planes were taking off, airline representatives said there was little to no effect on flights.

Southwest, which has suffered two other, unrelated technology issues this week, said its website and other internet-based tools were briefly disconnected but flights were not affected. United said there were no lingering issues early Thursday.

The disruptions occurred only days after many of the world's top websites went offline briefly due to a software disruption at Fastly, another major web services company. The company blamed the problem on a software bug triggered when a single customer changed a setting.

Brief internet service outages are not uncommon and are only rarely the result of hacking or other mischief. However, the recent outages have underscored how vital a small number of behind-the-scenes companies have become to running the internet.