The dark side of the cloud
Last week's storm-driven outage of Amazon Web Services and other recent high-profile foibles have, not surprisingly, re-sparked the "Is the cloud really worth it?" debate among enterprise IT professionals.
This is likely to be a regularly recurring thing, particularly when such events translate into prolonged downtime for popular consumer websites, drawing even more attention to the inherent "dangers" of doing business the cloud. (Pinterest's site went down, for crying out loud—who in the business world could be expected to work or otherwise function until that national emergency was resolved??)
Granted, a major outage of a cloud computing platform as broad-ranging as AWS should warrant the attention of CIOs and others responsible for their companies' IT operations, cloud or otherwise. But it's really not an either/or question. The issue isn't whether those organizations should continue leveraging the cloud for faster, more flexible and more cost-efficient software and computing power. Of course they should, because it's good business and it follows the tech-sector trend of more available bandwidth, more flexibility and more efficiency at less cost than the alternative of companies operating their own individual data centers. The variable is reliability, and that is where the attention of IT professionals should be focused right now.
IT pros who leverage the cloud should be insisting that their cloud service providers not only have the best in failsafe backup provisions in place, but also that their reporting process for outages is buttoned up. Much of the criticism surrounding Amazon's recent outage was focused on the fact that the company didn't alert its customers in a timely manner about the outage, and that customers had no way of knowing whether they were looking at a few minutes, a few hours or a few days of downtime.
Enterprises with mission-critical operations should be insisting not only on service-level guarantees, but also constant updates and time-to-restoral estimates during outages. For many, that may mean a move to a more expensive managed hosting plan with more stringent service-level and reporting provisions--but it's a cost that likely is worth it in the long term. It also may mean more of a reliance on hybrid platforms that leverage both cloud capabilities and their own premise-based architectures for in-case-of-emergency backup.
Since we all know that the tech sector, like nature, abhors a vacuum, expect the newest round of cloud computing entrants to be emphasizing reliability, service level guarantees and transparency with customers as factors that distinguish their offerings in the wake of the Amazon outage. And since Google is among them with its new App Engine offering, maybe the cloud sector will see some positive changes with regard to reliability.
Should enterprise IT professionals be rethinking the wisdom of the cloud? Only if they weren't already thinking about reliability and weren't already holding their providers accountable for meeting their service level needs. --Jason Meyers