gracie’s blog

for those who are in the same business as I am, this would be such an interesting news. i wonder why this happened and how it happened…

IBM Outsourcing Crash in Denmark
Mette Ahorlu
April 22, 2008

IBM recently experienced what can only be described as a major service breakdown that affected a number of large outsourcing customers in Denmark. The crash was reported in the Danish press and on the www.denmark.dk Web site.

The Crash

A failure in a network switch impacted a number of IBM customers:

• Danske Bank, which has more than 1 million customers in Denmark, Finland, and Northern Ireland, was unable to use credit and debit cards.
• Arla, a Swedish-Danish dairy company, was unable to supply a range of outlets in Denmark.
• Carlsberg, the Danish brewery, was unable to accept orders.
• DSV, a Danish transport group, lost access to most of its fleet.
• Maersk Zealand, the world’s largest container transport company, lost access to some of its core global applications.
• Pharmacies in the Copenhagen area were unable to receive prescriptions from doctors (a mainly electronic process in Denmark).
• Citizens were unable to access Sundhed.dk, Denmark’s official health Web site.
• An unspecified number of Copenhagen Municipality’s 57,000 employees were unable to use their core applications.

The outage lasted for up to 18 hours for some customers.

The Impact

DSV has decided to cancel its outsourcing contract and take operations back in-house, while other companies are considering claiming damages from IBM, according to the Danish Ministry of Foreign Affairs.

The outage raises a number of questions. First, IBM did little to communicate with its customers or with the media, analysts, or the public during the breakdown. IBM claims this was because it does discuss specific customers. This may be the case, but in terms of damage-limitation the company was at best naïve. The breakdown did not just affect a single customer: IBM has a very strong outsourcing position in Denmark, with close to 30% market share, so the impact was wide-ranging.

IDC is surprised that the outage of one component can bring down a large part of a datacenter — and its backup center — for so long. Normally, operations would simply switch to the other center, but in this case that didn’t happen, and guaranteed availability couldn’t be delivered. IBM has some serious work to do in trying to prevent single points of failure in the future.

The question now is how customers should react. Should they conclude that IBM is not good at infrastructure outsourcing or (worse) that infrastructure outsourcing in itself is not safe? Some, like DSV, have already made their decision. On balance, however, IDC believes outsourcing will enable customers to save money and let them concentrate on what they do best. Should enterprises spend more time and money to develop contingency plans? In most cases, this is not a realistic option — and, anyway, it’s the outsourcer’s job to guarantee service continuity. Everyone would like 100% uptime, but this is not realistic, and nor is it economic. Making IT unaffordable is not an option. But neither is paralyzing large swathes of Danish industry. IBM simply cannot afford another crash like this.

The question is not about outsourcing being sufficiently stable. The question is if IT itself is sufficiently stable. Overall, it is. In Denmark, however, IBM appears to have an operational issue that it needs to fix before its reputation is seriously damaged in the Nordics, and maybe beyond. We do not expect the breakdown to trigger a wave of insourcing, but DSV’s decision to cancel its contract will do nothing to help improve IBM’s image.