X is recovering after a data center outage

X seems to finally be recovering from a data center outage that brought down the site for some users Thursday and caused lingering issues into Friday. According on the company’s developer platform page, a “site-wide outage” that began at 11AM PT on Thursday, May 22, had “been resolved” as of 10:35 AM PT Friday morning.

The developer site notes that X is still experiencing “degraded performance” of some of its login features. The company has yet to officially comment on the ongoing technical problems since an update Thursday afternoon, when the company said that a data center outage was causing “performance issues” for some users.

At the time, reports on , which tracks online service outages, spiked as users reported issues accessing direct messages and other features. While the company hasn’t elaborated on the cause of the prolonged outage, the timing lines up with a reported fire at an X data center in Oregon on Thursday. According to , firefighters responded to a fire at a data center leased by X near Portland, Oregon at 10:21AM PT on Thursday. The extent of the damage is unclear, but the fire crews were reportedly on-scene for several hours. Batteries were apparently a contributing factor to the blaze.

X hasn’t responded to questions about the fire or the data center outage it disclosed. However, this wouldn’t be the first data center-related headache X has faced. Shortly after Elon Musk took over the company in 2022, he insisted on moving the company’s servers out of a facility in California to a space in Oregon in a bid to save money. And while Twitter engineers had insisted the process would take months, Musk insisted on moving them in a matter of weeks, in an incident Musk’s biographer.

While Musk was able to accomplish his goal of quickly relocating the servers, his haphazard approach to the move resulted in months of technical issues for the company and by the Federal Trade Commission.

Update, May 23, 2025, 12PM PT: This post has been changed to reflect X’s latest updates on the outage. It was previously updated multiple times, and that information is now included in the story above.