I’m going back a month here because ZDNet’s Phil Wainewright wrote a column whose underpinnings bear repeating. The message essentially is that
the cloud – with all its sexiness, innovation, allure and hype – is no place for amateurs. If you’re an enterprise-grade SaaS vendor, businesses are entrusting their critical operational data to you. If something runs afoul, it’s not as simple as reinstalling iTunes to see if your playlists are fixed. It typically means a major segment (or segments) of your business are rendered inoperative. And that, as any executive knows, boils down to bottom line impact.
When things go wrong – and they will, because that’s life and betting on 100% seamlessness is a fool’s errand – who do you want in charge of making things right? A pro or amateur?
Wainewright provides a few examples of amateur error handling in the cloud:
The boss of Air New Zealand has given us a convenient term for companies that can’t get to grips with the realities of delivering computing as a service: “Amateurs”. His reported comments were addressed to IBM, which failed to restore operations at a mainframe data center in a responsive enough fashion after a major outage on Sunday:
“In my 30-year working career,” he reportedly emailed the hapless vendor, “I am struggling to recall a time where I have seen a supplier so slow to react to a catastrophic system failure such as this and so unwilling to accept responsibility and apologise to its client and its client’s customers.”
And
T-Mobile is another reputable company left looking amateurish today after the catastrophic loss last week of all user data stored on its Sidekick service. But the real amateurs behind this story appear to be Microsoft and Hitachi, who are believed implicated in a server failure that took out both the production and backup databases on the storage network where Sidekick data is stored.
When it gets down to bare metal, you very quickly learn if you’re dealing with a company who’s playing in a sandbox without the proper toys. And no matter what you were sold – uptime, ROI, cost savings, integration, efficiency – all of it begins to pale when the strategic decision you made is proving to be a revenue bottleneck (or worse) because your chosen SaaS vendor doesn’t have the recovery chops it needs.
So that’s Wainewright’s amaetur. What’s his pro? Simply put, his retelling of how Workday handled their recent issue. How’d it go? In a word – professionally.
To read a contrasting story that shows how cloud outages get handled professionally, check out Michael Krigsman’s post last week about the recent 15-hour outage suffered by on-demand ERP provider Workday. Here, too, a network storage device caused a total meltdown, shutting itself down when it detected a corrupted node in a backup disk. Workday avoided Sidekick’s fate by invoking its disaster recovery plan. It avoided IBM’s fate by acting rapidly and going out of its way to keep its customers informed.
The smart SaaS vendor is beginning to understand what Workday already knows: that saying you’re a SaaS player doesn’t mean you repackage/re-host the stuff you already do behind a single sign-on screen. There are SaaS business models and scalability concerns that you need to confront. There’s deployment and integration strategies. Data security. Tenancy questions.
If you don’t look at each and every one of these things — if you instead opt for the lightweight, low-cost entry into the hot SaaS market — you’ll not only fail, but you’ll bring down the cloud pros who have taken the time, hired talent, and spent necessary dollars to address these issues for their customers. And that, in turn, undermines even the most professional cloud vendor’s efforts.
