Feb 25, 2009 (12:02 PM EST)
Coghead Failure Highlights Risks Of Cloud Computing
Read the Original Article at InformationWeek
Govind Davis, a Coghead customer, didn't see its sudden closing coming. Enticed by the cloud computing vendor's low-cost service, Davis built some of his customers' applications using Coghead's online database. Now his company, MCF Technology, has nine weeks to extract data from Coghead and build new applications for it, some of which will run on Intuit's QuickBase, also in the cloud.
Davis is still a huge proponent of cloud computing -- but a wiser one. "We'll be a little more wary with startups, as we realize the risk of going with one can be real," Davis said.
On Tuesday, a few days after Coghead said it was shutting down, a more widely known cloud vendor -- Google -- had a brief outage of its Gmail service because of a data center problem. Undaunted, Google announced that same day that it will start charging for Google App Engine, its Web-based application platform (or platform as a service).
No one should expect cloud computing to be infallible; software and servers never are, no matter where they reside. But Coghead's closure lends sobriety to the excitement surrounding cloud computing, which has attracted entrepreneurs big and small and fueled a new wave of venture capital in recent years.
Coghead's sudden closure also highlights the biggest gotcha of a platform as a service, software as a service, or anything else in the cloud genre: The vendor controls the systems and software. Coghead has told its customers their data needs to be out of its systems by April 30. If a traditional software vendor suddenly went belly up, at least its customers wouldn't face the pressure of a nine-week timeframe to find new software.
Choosing a stable vendor can reduce some of the risks of cloud computing. But even Google acknowledges that the migration path off the Google App Engine, should its customers want one, needs some work. "This is an issue, and we know it's not as easy as it should be," said product manager Pete Koomen.
Google offers an open source data-uploading tool, and to ease migration, will soon offer an open source downloading tool, Koomen said. Google supports the Python-based Django Web development platform, and if customers write their apps to conform with it, data portability is going to be the easiest, he added. Data portability "is something the team is very interested in and will make a lot of progress on going forward," Koomen said.
With Google's first-time payment structure for the previously free-with-quotas Google App Engine, customers pay 10 cents per CPU core hour, 10 cents per gigabyte of traffic in, 12 cents per gigabyte of traffic out, 15 cents per gigabyte of data stored per month, and $1 per 10,000 e-mails sent.
It's that type of low-cost payment structure that attracted Davis to Coghead. He's been an Intuit QuickBase customer for a while and still is, but Coghead was a little cheaper, starting at about $10 a month per seat.
"There's all kinds of reasons why a platform as a service is valuable, and that doesn't change because Coghead fell apart," Davis said. "If people want a low-cost way to deploy apps all over the world, then a Web-based database is the absolute best choice."
QuickBase accounted for just a portion of Intuit's $3.1 billion in revenue last year, but the company has customers big and small for it, said Intuit VP Bill Lucchini, including Affinity Health Care, with 15,000 seats, and another large company, with 50,000 seats. When it comes to SaaS, PaaS, or anything related, "a proven vendor is your first line of defense here," reasoned Lucchini. "You must choose vendors that are stable."
If a customer wanted to move its data from QuickBase to something else, Lucchini said, they can use QuickBase Desktop, which he described as providing a Microsoft Access replica of a customer's data. Some customers currently use it for backup or to work offline; it synchronizes with QuickBase when back online, he said.
The promise of cloud computing has launched many PaaS startups: Caspio, LongJump, TeamDesk, Wolf Frameworks, and PerfectForms are among them. It would be unfair to link them to Coghead just because they're startups, as some of them may grow into huge success stories. Still, others will not. And with any cloud computing effort, it's up to users to have an emergency exit strategy in place. Just in case.