TCS has already been tested
That level of service is important, because often different carriers or network service providers may offer services on the same physical cable, Neemuchwala said. "As a part of our network design we focused more on what is the physical fiber and we make sure for the primary and secondary links they are not on the same fiber. So we take two different routes altogether," he said.Neemuchwala said for the most part customers did not notice the latency, although as a part of TCS's business continuity plan, the outsourcer did notify customers of the switchover. "We informed those who had a primary link outage that we switched them to the secondary," he said. The automatic rerouting typically took five or 10 minutes to execute, although in one instance a carrier took three hours to do it circuit by circuit. That manual effort was meant to insure that the huge amount of traffic that needed to be rerouted would not all end up on the same secondary circuit. As a part of its business continuity plan, TCS builds in multiple layers of redundancy into its infrastructure as well as with its service personnel. "And we insure there is no single point of failure," he said. As a global outsourcer, the undersea cable cuts were not the first time TCS's business continuity plans were tested. Over the past several years, the company has used its business continuity plans to avert business disruptions in China during the SARS health crisis, in its headquarters in Mumbai when heavy rains washed out connectivity, and in the Eastern United States during a widespread power outage over the last several years.
Of the 200 different WAN links that TCS uses, only 15 links were affected. Still, traffic rerouted to a secondary link did see increased latency because the secondary link only provides half the capacity of the primary link.