Unspecified “power issues” at one of the data centers hosting infrastructure of San Francisco-based Salesforce.com, a popular cloud-based customer-relations-management service, brought down seven instances of the company’s cloud early on Tuesday morning.

This was a second outage for the company in two weeks. In late June, one North American instance of the company’s cloud went down as a result of an issue in its storage infrastructure, according to news reports.

Service disruptions were first detected around 12:30am PDT, according to the company’s status dashboard. In an update posted around 2:45am, the company said the power issues had been fixed.

“Power issues were detected but our technician onsite has confirmed this has been fixed,” the update read. “We are currently working to restore the service.”

The names of three of the instances that went down (NA1, NA5 and NA6) suggest that they were hosted in North America. The other four (CS0, CS3, CS1 and CS12) are part of the Sandbox infrastructure, where customers can deploy their cloud-based environments for development, testing and training.

Most functionality for the majority of affected instances had been resolved by about 11:30am. Around 12:10pm, the company posted an update saying the outage had been completely resolved, and all services had been restored.

In the process of restoring functionality, the company’s team attempted to fail one of the affected instances to a disaster-recovery data center. The techs quickly changed their mind about the failover, estimating that they could restore functionality within the primary data center supporting this instance in the same amount of time and with lower risk.

There was another major outage reported Tuesday. A Level 3 data center in London was knocked out by an uninterruptible power supply (UPS) fault.