Equinix data center in Paris that experienced a significant cooling system failure earlier this week, causing at least one customer to temporarily stop all operations at the location, has been returned to normal operation. Cooling systems at the colocation facility have been repaired, "all primary capacity is back online and the reserve system that had been helping to support the site is in its usual standby condition," Equinix France General Manager Michel Brignano wrote in an email to a French Network Operators Group mailing list around 1 p.m. PDT on Friday.
On Wednesday evening (European time), multiple chillers that supported second floor of the Paris 2 Internet Business Exchange facility, impacting temperature on that floor, indirectly affecting the ground floor as well, according to Brignano's earlier report.

Hallway at an Equinix data center in San Jose, Calif.
The standby chiller system did not start in time to absorb the load created by the primary chiller failure, specific reasons for which were not known or disclosed as of Friday. "There appear to have been component/subsystem failures in at least two of the three primary chiller systems supporting the second floor," Brignano wrote. "At this time, we cannot say definitively whether the failures were related or not."
One of the facility's tenants - Web hosting provider Gandi IWI - reported on its blog around 11:30 p.m. CEST on Wednesday that the air conditioning failure had impacted about one quarter of its hosting services, all of its site maker services and one of its DNS clusters. By about 12:40 a.m. on Thursday, as technicians were working on a resolution and temperature in the data center began to lower, Gandi was able to partially reactivate its services and reported at 8:15 a.m. that the incident had been fully resolved.
Engineers and supplier support personnel, dispatched to the site Wednesday evening, returned the facility to target temperature early Thursday morning by bringing the standby system online and using some primary cooling capacity made available by partial repairs done to one of the failed systems.
In his Friday update, Brignano wrote that the facility's operators were preparing for forecasted high temperatures during the upcoming weekend and would have "increased staff presence on site to monitor operations closely. Failure analysis and corrective measures is work that is ongoing and will be communicated to customers."