The Downside of Good Things

Recently I returned from a trip abroad and switched on my PC. After only 8 days away the number of components needing updating (Windows / Office / Norton / Adobe / HP / Safari / iTunes..) combined to break my personal record. It was 51 minutes from the time I switched on my PC to the time I was first able to do something productive.

It was only 2 years ago that I went past the 30 minute mark. It can’t be long before the hour is reached. Has anyone else got there before me?

While drinking my second coffee, I wondered if this was evidence that moving to a thin-client cloud-base would be more productive. Then I remembered last December when staying down in Dartmoor. The internet was running at dial-up speeds and fell over every 20 minutes or so for 4 days.

Looking at total system availability over the last year, I have had more time lost to network problems than to hardware or software problems. I am not alone. Computer Weekly reported recently that power outages and communications failures are the biggest causes of business disruption, contributing 3 times more incidents that hardware outages.

How much of a break will this have on the development of Cloud Computing? Adding an extra 9 to 99.99 availability at the data centre end is all well and good, but what is the benefit of that, if the weakest link is elsewhere.

One of the consequences of the visibility of blogging I have found is sometimes that I get approached by readers who would not have found me any other way.

The research department of a US law firm found my earlier blog entry on energy security. In the aftermath of the terrible devastation in Japan, they called to ask me my opinion on one question:

‘If the Japanese infrastructure was largely cloud-based would Japan have been more or less resilient in the aftermath of the earthquake and tsunami’.

As good systems people, we know that all systems to be effective need redundancy. We now have the sight that production of cars and electronic items are being hampered because of a shortage of components from Japan.

As a discipline, we in ICT have now contributed to making the global supply chains so efficient it is quite extraordinary, even looking back 10-20 years.

Yet, have we sought efficiency at the expense of resilience? Are we potentially creating the types of systemic risks that hit the financial sector in 2008?

Efficiency is a good thing. More efficiency must be better than less efficiency. However, reductio ad absurdum, we get some strange effects. Markets are efficient. Efficiency is a good thing; therefore each sheet of paper used by the civil service should be separately tendered for by competitive tender... to get the best deal for the tax payer. This is obvious nonsense.

Just when Japan needs economic activity to contribute to the recovery, that activity is being hampered by past efficiencies.

In Japan, some areas have lost both power and parts of the communication infrastructure.

The UK gets only minor tremors, but the flooding in Gloucestershire nearly hit the National Grid back in 2007. The Pitt review is interesting reading on that incident!

In responding to the American question, I came to a personal judgement. Within a decade, a major organisation may well go into administration because it cannot access its operational data in the cloud after an outage of power or telecoms.

So, the interesting challenge for public, private or hybrid clouds I suspect is ‘who manages the business risk?’

My concern is that IT cost efficiency through Cloud-based Computing badly handled may contribute to increased systemic business risk.

My experience of the Japanese is that they are that they are a proud people and they will rebuild. They will also learn the lessons of this disaster. They will adapt. While we in the West spend our energy trying to pin the blame on someone, they fix the systems.

For cloud computing to achieve its maximum potential, we need efficiency and resilience. The Service levels that matter are end to end and as seen by users. How will cloud suppliers deliver that?

I’m getting deeply unimpressed by the huge size of some updates and the productive time lost, but I’m not giving up my data yet!

So the question I want to pose is this:

‘For Cloud Computing to mature in the enterprise space, what availability of power and comms is needed before business risk is acceptable?’

If you can answer that, then when that will be deliverable?

About the author
Chris is a technology and policy futurologist. Chris has been in the IT industry since 1980. His roles have spanned Honeywell, ICL, HP, Microsoft and Capgemini. He is a Fellow of the BCS and a Fellow of the RSA.

See all posts by Chris Yapp
April 2018

Search this blog