Today's turbulence explained
Not how we wanted today to go. At approximately 9:00am PDT, we discovered that Twitter was inaccessible for all web users, and mobile clients were not showing new Tweets. We immediately began to investigate the issue and found that there was a cascading bug in one of our infrastructure components.
Context & Ripple Effects
This is the third time Twitter has had to publish an outage explainer on its own blog — after the August 2009 'Adventure Continues' post and the June 2010 'What's Happening with Twitter?' post — but the company around the failure has changed shape. The same-day @twittercomms confirmation of a cascaded bug landed in a week when Twitter was simultaneously pushing promoted tweets into Latin America and western Europe on its way to advertising products across 50 countries by year-end.
The breadth of pickup — TechCrunch, AllThingsD, CNET, The Verge, GigaOM, Computerworld and the LA Times all carried the story within hours — is itself the signal: a site that was once a hobbyist service is now treated as daily infrastructure, so a morning of downtime is national news rather than a niche grievance.
First-order effects
- Web users worldwide lost access and mobile clients stopped showing new Tweets from roughly 9:00am PDT, making this a full-platform failure rather than a partial degradation.
- The outage hit mid-rollout of Twitter's international advertising push, putting paid-product credibility in front of new advertiser markets on the same day the service was dark.
Second-order effects
- Enterprise data buyers are exposed too: days earlier Twitter granted Salesforce's Radian6 access to its full tweet firehose, and commercial firehose contracts make uptime a contractual concern, not just a user-experience one.
- Rivals and partners watching the ad ramp now have a concrete argument that reliability, not just reach, will decide whether brands shift budget onto the platform.
Third-order effects
- If cascading-failure outages keep recurring as the service scales, Twitter's engineering narrative shifts from the Fail Whale era of capacity excuses to architectural resilience as a published, brand-level commitment — the difference between a free service that fails and a revenue platform that cannot afford to.
The trend: Consumer social platforms crossing into advertising and enterprise-data businesses are converting uptime from an operational detail into a commercial promise their revenue plans depend on.