OVH has a hunch about what caused the fire that destroyed its data center

cloud
(Image credit: Shutterstock / Blackboard)

The CEO of OVH has hinted that shoddy maintenance work on a couple of UPSes could have ignited the blaze that destroyed one of the web hosting provider’s four data centers and knocked major websites and services offline.

OVH, which is Europe’s largest hosting provider and the third-largest globally, recently suffered a catastrophic fire at its facilities in Strasbourg, France. 

In a video message to apologize to customers and give an update on efforts to get its data centers back online, company chief Octave Klaba pointed to the recently serviced UPSes as possible culprits for the incident, as he explained the reasons for the blaze, towards the end of the video. 

Investigations

Klaba confirmed that the fire completely destroyed the SBG2 data center, as had been reported, adding that four rooms of SBG1 that were close by were also lost in the inferno.

However, the other eight rooms of SBG1, along with the two other data centers, namely SBG3 and SBG4, are completely unharmed. 

As he recounted the events on the morning of the fire, Klaba said that by the time its staff responded to the fire alarms, the facility was already engulfed in smoke.

When the fire department responded to their hails, they also bought thermic cameras, which allowed them to see through the spoke and spot the two USPes that were on fire. Klaba pointed out that one of these particular UPSes had been repaired a day earlier, and was pushed back into active service in the afternoon before the day of the fire. 

He also pointed out that the gutted SBG2 data center has over 300 security cameras and the team is working to retrieve and comb through to determine the culpability for the event. 

At the same Klaba assured customers that the team is working hard to restart the remaining data centers after thoroughly verifying the cooling systems and the power supply systems, which is a lot of work. However, they still hope to be able to get some of the systems back online by next week.

Via: The Register

Mayank Sharma

With almost two decades of writing and reporting on Linux, Mayank Sharma would like everyone to think he’s TechRadar Pro’s expert on the topic. Of course, he’s just as interested in other computing topics, particularly cybersecurity, cloud, containers, and coding.