CenturyLink Routing Issue Led To Outages On Hulu, Steam, Discord, More
A CenturyLink BGP routing mistake has led to a ripple effect across the Internet that led to outages for numerous Internet-connected services such as Cloudflare, Amazon, Garmin, Steam, Discord, Blizzard, and many more.
These outages started at approximately 6 AM EST, when customers began reporting a wide-scale outage in the USA affecting CenturyLink services.
When performing searches on Twitter, there was a sudden influx of complaints about poor performance or outages on numerous connected services such as Blizzard, Steam, Discord, Roblox, Cloudflare, Hulu, Slink, Reddit, Amazon AWS, and many more.
CenturyLink states that their Level3 CA3 data center is causing this outage and are investigating the issue.
“Our technical teams are investigating an issue affecting some services in the CA3 data center. Ensuring the reliability of our services is our top priority. We will continue to provide status updates as this incident progresses. If you need further support, please contact us at [email protected],” CenturyLink’s status page states.
This outage has since been resolved, and services are slowly recovering, with some areas taking longer than others.
These issues were also likely related to the NameCheap outage we reported earlier.
BGP Routing issue caused outages
According to numerous reports from affected customers, today’s problems were caused by a BGP routing issue at CenturyLink where they were not correctly routing portions of the Internet.
For the Internet to work, Internet service providers, data centers, and network providers advertise via the BGP routing protocol the IP addresses that they route and manage.
As this is mostly a trust-based system, when a large ISP starts advertising routes for IP address ranges that they do not manage, it causes worldwide outages and performance issues.
Based on customers’ reports, CenturyLink appears to have made a mistake in the BGP routing, which led to today’s wide-scale issues.
Cloudhelix, a cloud hosting provider, also experienced issues today and state that CenturyLink has confirmed it was a BGP issue.
“CenturyLink have confirmed a routing issue in their network was preventing BGP sessions from establishing correctly.
They have now rectified the issue and their network is stabilising as the updated configuration propagates to affected devices.
The CenturyLink connection at Cloudhelix’s facility in Equinix LD6 continues to be affected. We will issue another update in due course.”
CenturyLink has not updated their status message yet or publicly confirmed that it was a BGP routing issue.
This is a developing story.