Gmail follies and Google’s enterprise pitch

Last week the No. 1 objection to cloud computing raised its ugly head again: Gmail went into Gfail mode, leaving users stranded for over 90 minutes. By our own Robert X. Cringely’s calculations, last Tuesday marked “the ninth notable Gmail outage in little more than a year.”

These incidents come at an unfortunate time for Google. A couple of weeks ago, I spoke with three product managers, plus Dave Girouard, Google’s president of enterprise, about Google Apps and their immediate future in business. I was surprised at how far things had advanced.

[ Robert X. Cringely offers his take on the Gmail outage in “Gmail Gfails, Internet survives again.” ]

Led by Gmail, Google Apps has 1.75 million users and, in its Premier Edition, “hundreds of thousands” of paid users at $50 per seat per year. Google was even able to point to a large biotech company, Genentech, that had rolled all its employees onto Google Apps. Integration with BlackBerry Enterprise Server has been achieved, and administrators can now enforce e-mail retention policies and delegation. Customers can even opt for the Gmail mail server while allowing users to keep Outlook as a front end. And of course, Google Gears now allows you to use Google Apps offline.

All very impressive — but I had to bring up the availability question. Unlike downtime on internal enterprise systems, every outage is very public. So while on-premise systems may go south more often (I won’t go into detail about a rather painful e-mail migration here), Google’s claim of 99.9 percent uptime still yields close to 9 hours of high-profile downtime per year. So I got assurances about redundant “active/active” replication to protect against data loss, SAS 70 compliance, and the elimination of planned maintenance downtime (which normally plagues in-house Exchange implementations).

I’m sure that’s all true. But what does Google offer customers who experience problems? A service credit of up to 15 days. At $50 per seat, that’s chump change.

As was clear from an affable, academic-sounding blog post by Ben Treynor, Gmail’s “site reliability czar,” Google and enterprise customers are still talking different languages about last week’s outage. Can you imagine your reaction if, after an e-mail outage in the middle of your workday, you got a chatty message from your e-mail admin about some problem with Exchange?

I had a similar reaction to the newly announced Virtual Private Cloud offering from Amazon. Yes, the company has added some security features enterprises desire. But scroll down to the bottom of the FAQ, and in answer to the question “Does Amazon VPC have a Service Level Agreement (SLA)?” you’ll see a two-word answer: “Not currently.”

And when you think about it — how could they? The liability would be ridiculously large if Amazon or Google started offering guarantees that came near the value of some unforeseeable disaster. Until that elephant in the living room is dealt with, few enterprise IT folks will take cloud services seriously.