While British Airways counts the cost of its May bank holiday system outage – £80m and growing so far – much of the datacentre industry has listened to the supposed cause of “human error” and thought, “Are you sure?”
Willie Walsh, CEO of BA parent IAG, said the problem was caused by an engineer disconnecting the power supply to one of its datacentres, before reinstating the power incorrectly, leading to a power surge that damaged equipment.
Most experts that Computer Weekly talked to felt this scenario was – or should be – near impossible in a modern, well-designed datacentre. Either there is more to the issue than has yet been revealed, or it reflects badly on the datacentre design or set-up at BA.
It smacks of an over-reliance on legacy systems – BA is believed to still have software written decades ago for ageing mainframes. Modernity costs money – sometimes it’s seen to be better to maintain the creaking status quo, at least until it becomes too expensive not to.
The UK’s big retail banks are notoriously reliant on extraordinarily complex legacy systems that would cost billions to replace or modernise. So far, only Royal Bank of Scotland has suffered the sort of catastrophic outage on a similar scale to BA – but you can bet that CIOs at the other high-street banks knew it could have been them.
Not all legacy systems are bad, of course. But maintaining legacy that is clearly a long-term hindrance to any organisation is a management decision, not a technical one. It cannot be long before large, well-established companies start to lose market share and fail to compete sufficiently because of their legacy systems – which will then become the root cause of the collapse of a business, not just the cause of an IT outage.
CEOs and CIOs at the sort of large organisations likely to be at risk rarely stay in the job more than three to five years these days – not long enough to have the courage or mandate to take on such an enormous task as overhauling all that legacy complexity. In some cases, it’s simply too big a problem for anyone to take on.
But many of the industries least affected by digital disruption so far are also those with the biggest legacy investments – banking, manufacturing, insurance, for example. When digital revolutionises those sectors – and it will – the legacy laggards will be in serious trouble.
Companies die – that’s inevitable. The FTSE100 in 1990 looks very different to today. But increasingly the factor that determines success or failure will be the managed withdrawal of the sort of legacy systems that remain all too common.