Oracle data-center outage cripples U.S. TikTok posting, creators report lags

TikTok’s U.S. service was disrupted on March 3 after an issue at Oracle’s Ashburn, Virginia data center, leaving creators across the country unable to post reliably and many users experiencing timeouts and slow performance. TikTok USDS, the U.S. joint venture that now runs the app, posted on X that “an issue with an Oracle data center is impacting some parts of the TikTok U.S. user experience” and warned that “Creators may temporarily experience lags in posting content while Oracle works to resolve the issue.”

Oracle acknowledged a service disruption on its System Status page and in posts on X. The company’s status message, as quoted by 9to5Mac, said, “We are continuing to investigate an issue with Oracle Cloud Network Infrastructure in the US East (Ashburn) region. Oracle engineers are currently taking steps to improve network stability in efforts to enable further progress toward impact mitigation.” Oracle also told users it was “working quickly to restore normal service operations,” according to outlet reporting.

User complaints recorded on Downdetector showed clear spikes during the day, though outlets differ on the precise timeline. Oracle’s System Status and 9to5Mac reported the incident began around 8:24 a.m. Eastern, and TechCrunch said user reports indicated the problem was ongoing before 9 a.m. Eastern. The Verge reported Downdetector activity spiking around 1 p.m. Eastern, while the Times of India said reports climbed just before 1 p.m., spiked again near 3:30 p.m., and that “thousands” of complaints had poured in by evening. Gizmodo described problems trickling in Tuesday morning, peaking later, and beginning to taper into Wednesday.

The outage is the latest operational headache since the January deal that shifted majority control of TikTok’s U.S. operations away from ByteDance to a new investor group that includes Oracle, Silver Lake, and others. TechCrunch and the Times of India note Oracle is part of the investor group that owns roughly 80 percent of TikTok USDS, and that Oracle has been providing cloud services and managing TikTok user data both before and after the ownership change.

This is at least the second Oracle-related disruption affecting TikTok U.S. since the sale. Outlets disagree on the timing of the prior incident, with some placing it in late January and others in February, but reporting consistently linked that earlier outage to an Oracle data-center power or weather problem. The Times of India and The Verge reported that the earlier disruption had interfered with TikTok’s recommendation algorithm and left some creators seeing zero views on new uploads, a failure that sparked conspiracy theories about the platform’s new owners. Gizmodo reported that TikTok denied deliberate censorship at the time and said the denial appeared genuine.

Oracle had not identified a root cause for the March 3 incident at the time of reporting, TechCrunch said, while 9to5Mac noted the company’s System Status page had published more than a dozen updates as engineers worked on mitigation. Beyond disrupted posting and higher latency for creators and viewers, the outage raises broader questions about operational concentration: a single cloud-operator problem at a major Ashburn facility can ripple through a platform with tens of millions of U.S. users and a nascent ownership structure that ties the cloud provider to platform governance.

Reporters following this story say the next steps are to secure Oracle’s archived status timestamps, precise Downdetector counts, and any internal TikTok metrics on failed uploads and recommendation anomalies to resolve discrepancies in the timeline and to determine whether other cloud customers were similarly affected.