SAN FRANCISCO — Salesforce Inc. suffered a widespread global service outage Wednesday, the second day of its flagship Dreamforce conference in San Francisco, leaving customers across multiple continents unable to access core parts of its cloud software platform at an unusually inconvenient moment for the company.
According to Salesforce’s own status page incident tracker, the disruption began around 7:50 a.m. UTC, or roughly 3:50 a.m. Eastern time. The company said customers across all three of its operating regions could experience severe delays, intermittent errors, or a complete inability to access some services. “During a service disruption, end users can’t access the service,” Salesforce said in an update posted to its status page as the incident unfolded.
The outage’s reach extended well beyond the United States. Reports of disruption affected hundreds of Salesforce instances across markets including the United States, United Kingdom, Germany, France, India and Japan, according to tracking of the incident. Independent monitoring services logged a sharp spike in user complaints, with one service reporting that 69% of submitted issues related to website access, 16% to app functionality, and 15% to login problems.
Salesforce’s engineering team initially explored restarting affected systems as a potential fix, but that approach did not resolve the underlying problem. In a later status update, the company said it was “no longer pursuing restarts as a path to remediation,” adding that “customers continue to experience severe delays, intermittent errors, and inability to access some services and support case creation.” Salesforce said it would provide a further update within 30 minutes or sooner if new information became available.
Salesforce’s investigation ultimately traced the disruption to an internal login service. According to the company, incoming requests were stalling while waiting for a response from that service, a bottleneck that consumed available server resources and cascaded into the broader access problems customers experienced across the platform. By 10:56 a.m. UTC, roughly 6:56 a.m. Eastern time, Salesforce said its engineering team had validated a fix on a test instance and begun rolling it out across all affected regions.
The disruption’s timing drew particular attention given its overlap with Dreamforce, Salesforce’s flagship annual conference, which runs from September 15 through 18 in San Francisco. The event is expected to draw more than 40,000 in-person attendees, with more than 200,000 additional people registered to participate online, and features more than 400 sessions this year centered on Salesforce’s push into what the company calls the “Agentic Enterprise,” its broader strategy around AI-driven business software agents.
That AI push has included a newly announced partnership with Anthropic, under which Salesforce is embedding customer data and workflow tools into Anthropic’s Claude chatbot through a plugin aimed specifically at sales teams. The partnership was announced alongside Salesforce’s fiscal second-quarter earnings release, in which the company reported revenue of $11.35 billion, up 11% year-over-year, and raised its full-year revenue guidance by $200 million.
Salesforce shares were already under pressure heading into Wednesday’s outage. The stock closed Tuesday at $255.65, down 1.46% from Monday’s close of $259.43, a pullback that followed a sharp 4.73% rally on Monday tied to anticipation ahead of Dreamforce. Shares fell a further roughly 0.5% in Wednesday premarket trading, changing hands around $254.40. Because Tuesday’s decline occurred before the outage began, it cannot be directly attributed to the disruption, though the incident adds a fresh factor for investors to weigh as they assess the stock’s performance through the remainder of the week. Broader index futures showed a mixed picture Wednesday morning, with contracts on the S&P 500 and the Nasdaq Composite both posting modest gains even as Salesforce shares slipped.
Independent status-tracking services showed the scale of the disruption extending across a substantial share of Salesforce’s infrastructure. One tracker identified confirmed issues affecting more than 1,030 individual system components in North America alone, spanning Salesforce’s core service across multiple data center clusters. Separate monitoring logged the outage’s overall duration at more than four hours, with additional shorter incidents flaring up later in the day, according to historical incident data tracked for the service.
Customers experiencing access problems were directed by outside monitoring services to consult Salesforce’s own Trust Status page directly to determine whether their specific instance of the platform was affected, rather than relying solely on third-party outage trackers, which can occasionally lag behind or imprecisely characterize the true scope of an evolving technical incident.
Wednesday’s outage adds to a recurring pattern of high-profile disruptions affecting major cloud software providers over the past year, incidents that have increasingly drawn scrutiny given how deeply businesses across industries now depend on continuous access to cloud-hosted customer relationship management and sales tools for their day-to-day operations. For Salesforce specifically, the disruption’s timing alongside Dreamforce, an event explicitly designed to showcase the reliability and capability of its platform to tens of thousands of customers and partners, added a layer of irony that was not lost on observers tracking the incident as it unfolded.
With Salesforce having confirmed a validated fix was being deployed across affected regions by late Wednesday morning Eastern time, the company’s engineering team is likely to face continued scrutiny in the coming days over both the root cause of the login service bottleneck and whether additional safeguards will be put in place to prevent a similar disruption during future high-profile events on the company’s calendar.
You must be logged in to post a comment Login