A DR plan reduced outages and helped maintain uptime for Ericsson

Business case: For almost a year, Tamal Chakrovarty, CIO at Ericsson, fought tooth-and-nail with unit heads and business users to arrive at a consensus. And failed. The business' contention was simple: reduce the number of infrastructure and service outages the enterprise was facing. Chakravorty agreed to bring it down from double-digit outage days a year to three days, but the business wanted the number to go down to a couple of hours. At no extra investment. Failing to arrive at a consensus, one morning three months back, Chakravorty walked into a meeting and announced that his team was going ahead with a DR implementation. It was a unilateral decision that would have been hard to get

away with in good times and the slowdown made it harder. But the project  would enable Ericsson to reduce the number of days of outage to three a year. "If we had continued to strive for consensus, the DR project, in whatever form, would not have commenced. I would have bled all my resources to chase zero-day downtime and would have still been apologizing to the business every time an outage occurred," says Chakravorty. His railroading approach, while unusual, was essential if Ericsson wanted to tame network blackouts, which added up to about ten days every year, creating revenue losses of about a crore a year. If the global shared service centers and network operating centers blacked out, various services such as finance and other payment modes that were used by telecom operators, were disrupted, too, resulting in losses. "Business was furious. We understood their problem but our SLA contract with our outsourcing partners, which was signed about four years ago, clearly stated that double-digit days' worth of outage was acceptable. Going by the book, we couldn't have done anything about it.

Project: Nevertheless, IT decided to work around the problem and bring in a business SLA. That's where the problem began," recalls Chakravorty. The business returned with a DR plan that blueprinted zero-day outage. Based on the amount of bandwidth at various locations and the infrastructure and manpower that could be tapped with the current investments, Chakravorty reckoned they could trim service outages for Ericsson India and Sri Lanka to three days. "On paper, we still have the double-digit figure in the  contract. But with our softhandling of the issue and vendor management practices, we managed to get our outsourcing vendors to offer us more value-add for our investment. However, we still can't penalize our vendors for not maintaining blackout under three days," he says.

First steps: It's been three months since IT has put in place all the tools, technologies, processes and practices to keep its promise. And as much as the business didn't like the idea of not having zero-day downtime, Chakravorty's railroading approach opened its eyes to the efforts to curtail blackouts. "Users are beginning to realize the amount of effort required by IT to maintain uptime of critical systems. In fact, the business has slowly begun to release some more funds to invigorate the DR backbone," he says. The business now understands better, says Chakravorty, what they can and cannot get for every investment they make in IT. With business' increasing openness to funding the DR project, Chakravorty plans to plug more loopholes. "If  investment comes in, we can gradually take the three-day outage further south," he says.

Benefits: And the benefits of Chakravorty's approach keeps spreading. "The business now realizes that maintaining uptime is not as simple as maintaining a  home computer. It requires planning, investment and effort. The DR project that we forced on management has now started a chain reaction. Not only do they realize the extent to which they can make commitments to their clients, but it also opening up funds for more DR and business continuity projects," says  Chakravorty.

The Person Behind It

Tamal Chakravorty

Ericsson, CIO

Users are beginning to realize the amount of effort required by IT to maintain uptime of critical systems.