Gartner: Stop worrying and love the cloud, with all its outages and lock-in
Yes, there are risks. But you can address them without huge multi-cloud failover rigs
Stop worrying about the resilience of clouds, the possibility you'll be locked-in, or the systemic risks they pose – and instead learn to appreciate their benefits and limitations.
So said Alan Waite, a managing veep of analyst firm Gartner's Technical Professionals advisory service, at the org's IT Infrastructure, Operations & Cloud Strategies Conference 2023, in Sydney on Monday.
Waite said that concentration of cloud services in three big suppliers – AWS, Microsoft and Google for most of the world, Alibaba, Huawei and Tencent in China – does represent an unhelpful concentration of market power and operational risk.
But he said that conventional vendor management practices should suffice when dealing with the big clouds. And he told attendees that Gartner data on cloud resilience suggests that outages are becoming less frequent, and less severe. Most typically impact a small number of services, in a single region, for a couple of hours. The prospect of your operations – or, say, the global financial system – experiencing a major outage due to a colossal cloud failure are slim. And if such an incident did occur, an individual org's efforts to improve resilience would make little impact.
He therefore advised using each cloud's own resiliency features to the maximum extent possible, rather than contemplating active-active multi-cloud failover rigs.
While that means committing to your cloud-of-choice's way of getting stuff done, and therefore likely feeling locked in, Waite said every technology decision involves lock-in to some extent.
"If you've if you've worked in IT for any period of time, you've always been locked in in some way," he said. "Perhaps it was VMware in a datacenter. Perhaps it was Cisco for your network. You were happy to accept that lock-in because those vendors gave you fantastic capabilities that you could leverage and use.
"Cloud providers are exactly the same. There's no difference in lock-in to a cloud vendor. People accept lock-in understanding the risks that are involved with it. And they manage those risks. And they accept it because they get fantastic capabilities from their cloud provider."
On the resilience front, he advised using a second cloud for what he called a "lightweight alternative" to applications and services – a cut-down version that satisfies customers and regulators for the hopefully short duration of an outage. To illustrate the concept, he suggested retailers could create a lightweight version of their website based on static content and house it on a second cloud. If a retailer's main cloud falls over, the lightweight alternative would allow sales to continue – albeit in a less sophisticated fashion. Banks, he said, should consider what regulators define as a minimum required service and make their lightweight alternatives housed in second clouds capable of meeting that requirement, and not much more.
Such lightweight alternatives would sync data, so they can quickly be made operational.
- Amazon CEO says AWS staff now spending ‘much of their time’ optimizing customers’ clouds
- Cloud customers are wasting money by overprovisioning resources
- The world was promised 'cloud magic'. So much for that fairy tale
- VC's paper claims cost of cloud is twice as much as running on-premises. Let's have a look at that
Waite also advised users not to fear loss of control when adopting clouds. On-prem tech produces many metrics on system health that clouds don't provide, so the move to a cloud can make ops teams nervous.
The analyst urged orgs to trust in clouds in the same way individuals trust airlines to fly planes safely – and noted that long-distance driving is riskier than flying commercially.
Waite also touched on cloud cost optimization, explaining that the easiest route is cutting down on waste and overspend – which he rated the most common risk of a move to the cloud. ®