Опубликовано: 18.01.2010 Источник: www.datacenterdynamics.com
As a result of failure of a UPS module caused by a “power issue,” many customers of a Rackspace data center in London experienced downtime on Monday morning. More than 200 servers were down for several hours.
The outage comes less than one month after a networking issue that occurred in the process of bringing a new US facility online prevented many customer Web sites hosted in the company’s Dallas-Fort Worth data center from being reached for more than half of one hour. While the December outage was caused by a networking issue, Monday’s incident was caused by a failure in the facility’s power infrastructure.
“These interruptions were the result of a power issue caused by a UPS module failure on UPS B,” Rackspace stated on its Web site about the London outage. “The UPS then failed to transfer the load as expected. Chloride engineers have been notified and we will launch an investigation into what caused the UPS issues.”
Chloride is a UK manufacturer of UPS systems for data centers.
Rackspace did not specify how many customers were affected, but said that while more equipment had experienced problems, a particular group of about 220 servers took several hours to bring up as technicians had to “manually intervene.” In some cases, staff had to replace server power supplies and firewalls and reconfigure switches.
The London data center issue occurred around 9:19 a.m. and about two hours later, Rackspace posted an update, saying that power had been restored and its staff were in the process of bringing affected servers back online. |