Rackspace Goes Down. Again. Takes The Internet With It. Again.
Another day, another Rackspace outage. The hosting company had a complete and total failure today that took down a number of big sites on the Internet, including ours. This has been happening all too often in recent months, including downtime just last month.
Context & Ripple Effects
This is at least the third major Rackspace failure in the corpus's arc: a two-hour power outage at its Dallas data center in November 2007, a follow-up that showed how widely customer sites were hit (RackSpace Outage Hits Home), and now a complete hosting failure in December 2009 that TechCrunch says came on top of another outage just last month.
The pattern matters because Rackspace sells managed hosting on the promise that customers never have to think about infrastructure — and each repeat failure erodes exactly that pitch. The breadth of sites taken down today also makes the incident visible beyond the hosting industry, since readers experience it as 'the Internet' going down rather than one vendor.
First-order effects
- Rackspace's customers — including high-traffic publishers like TechCrunch itself — are offline entirely right now, with revenue, ad impressions, and audience trust lost for the duration of the failure.
- The company faces immediate service-credit obligations under hosting SLAs and a support burden from customers who were told reliability was the product they were buying.
Second-order effects
- Customers burned by back-to-back failures have a concrete reason to split workloads across providers or renegotiate contracts, shifting pricing power toward whoever can document better uptime.
- Competing managed-hosting and cloud providers gain a sales opening against Rackspace specifically on reliability claims, since a rival's outage is the cheapest proof point a competitor's sales team can ask for.
Third-order effects
- If single-vendor failures keep taking large swaths of the web offline simultaneously, redundancy across providers stops being an enterprise nicety and becomes a default expectation for any publisher or business running on hosted infrastructure.
- Repeat publicized outages push uptime from a marketing claim toward something buyers audit independently, raising the bar for what hosting vendors must prove to win and keep contracts.
The trend: Managed-hosting buyers are being pushed by repeated vendor-wide outages toward multi-provider redundancy, making documented uptime rather than promises of it the real competitive currency.