Short circuit fire at Airtel data centre in Malad, Mumbai

What happened

In late December 2011 a fire broke out at Airtel's data centre in Malad, a suburb of Mumbai, India. The fire was reportedly caused by a short circuit. It led to the shutdown of servers used to route calls in the region, causing a network outage that disrupted mobile services for many customers for several hours. Airtel said the outage affected its Western region as a result of the fire at its central point of presence in Malad.

Data services, especially to corporate customers in the region, were hit hardest. The site is the central point of presence for Airtel's Western region, where about 17 percent of the company's customers are located. The report also cited a 2008 short circuit at another Airtel data centre in the Mumbai area that caused a large fire and an outage of more than 12 hours.

Contributing factors

The reported cause was a short circuit. The source does not identify the failed equipment, the electrical system involved, or whether backup power was in use.

Lessons learned

1. Electrical distribution in critical data and telecoms halls needs regular inspection, thermal imaging and maintenance to find faults that can progress to a short circuit and fire.
2. Concentrating routing for a whole region in one central point of presence creates a single point of failure; operators should provide geographically separate failover.
3. Operators should review fire detection, suppression and isolation arrangements so that an electrical fire can be contained without shutting down the wider network, especially where a similar fault has happened before.

More lessons on detection and monitoring, from every incident on the register.

Source

datacenterdynamics.com