Gmail Failure: Reality Bites for Cloud Computing

Update: Gmail is back, after a two-hour outage. The company has not yet said how many people were affected, but twitter traffic from round the world reveals a major outage. That’s a pain for millions of ordinary users – but it could be a disaster for those planning to sell application services in the cloud.

The single biggest promise of cloud computing is risk reduction. Moving to hosted solutions is supposed to relieve IT professionals of the need to manage applications in-house, and allow them to replace costly investments with predictable regular payments.

At a single stroke, a substantial chunk of Google downtime has called those benefits into question.

The direct cost of the downtime may run to billions of dollars – at least according to one Twitter estimate: “25m users, 33 percent affected; average of $50 per hour lost productivity = $415m per hour economic cost…” So that’s $800 million then. But indirectly, this will subtly affect the shape of future IT.

Google has been a champion for services in the cloud – and Gmail has been its standard-bearer. Ordinary consumers have moved to the hosted mail service in droves, accessing emails from anywhere, and dumping the old emails which pile up in their home computer’s proprietary client.

More savvy business users have been aware of persistent stories of Google deleting accounts in error, and have used it carefully. They have synched to a mail reader, keeping a local copy, and rely on Gmail just for its superlative free spam filtering, and the ability to access anywhere. Google’s Gears-based offline feature just made it possible to do this automatically (at least for about six months or messages).

Small companies can use Google’s apps, offering the prospect of no tiresome mail servers to maintain at all.

And all this has been great publicity for those promoting Software as a Service (SaaS). It if works for email, then contact management, inventory and anything else can surely be put out on the web, where it will be handled by experts.

SaaS, like so many other ideas before, is a much-hyped panacea. There are already some cracks showing round it – a Gartner report last week warned that its benefits had been overstated, and warned that it might be cheaper for the first two years, but after that, it would be cheaper to invest in in-house expertise.

That kind of accounting argument can run for ever – and any side can make the figures of financial risk come out their way. Today, Gmail brings up the more immediate face of risk: catastrophic failure.

One failure of Gmail will give a lot of ammunition for IT professionals anxious to argue the case for in-house IT. We all know very well that in-house IT has its problems, but at least it is in-house, and we know who to chase when it goes wrong. Gmail users are faced with a blank uninformative statement from Google, and an official Google blog that still thinks nothing is wrong.

How seriously can something like this affect business? Google staff in the company’s London office reportedly had trouble dealing with the media and customers – because their email was down.