Grading our 2013 Cloud Predictions
At this time 12 months ago, we released our predictions for what changes in the market would be brought about by the maturing of cloud computing. Looking back on the year, we can now see that, while the promise of a maturing market was strong, maturity was by no means uniform and thus our predictions proved to be a mixed bag.
1. We’ll finally stop saying that everything is going cloud.
Grade: A. While the C-suite might still be preaching this as a long-term vision, we got real about what should and should not go to the cloud given its current maturity and capabilities. The guiding principles of architecture and economic model served as sufficient evidence that many traditional workloads have no business on the public cloud. And we started to see early signs of enterprises recognizing that the private cloud isn’t the new name for virtualization but is indeed a separate environment and not all apps in the data center are destined for this pool.
2. Cloud and mobile will become one.
Grade: A. As predicted, the dominant model for mobile applications was a backing in the cloud. Commercial mobile-back-ends-as-a-service are helping mobile applications elastically respond to mobile client engagements and shielding corporate data center applications from mobile traffic explosions. But we’d have to give a C to enterprises here who have been slow to embrace this synergy. We still see too many enterprises bypassing the MBaaS and trying to connect the mobile app to their web tier or right back to the back office. This burdens mobile applications unnecessarily and inhibits their agility while potentially pounding the backend as mobile becomes the primary system of engagement.
3. We’ll stop stressing about cloud SLAs.
Grade: C. While there is greater recognition that apps have to protect themselves, too many enterprises preferred to push for specific SLAs in their contract negotiations. While major cloud outages often prompt the desire for contracted SLAs, your own data center is no less susceptible than the clouds, so stop panicking over them. Since the clouds rarely negotiate their SLAs for individual customers, the end result for many an enterprise is a professional managed services contract (either with the services team of the cloud provider or a third party vendor who monitors and actively manages your workloads for you). See Bill Martorelli’s report on best practices in cloud SLA negotiations.
4. We’ll get real about cost modeling.
Grade: C. We definitely saw strong sales and growth for the cloud cost-monitoring tools like Cloudyn, CloudCruiser, Cloudability, Datapipe’s NewVem, and Rightscale’s Plan for Cloud, but we didn’t see them cross over to the majority of enterprise users of cloud platforms. And inquiries about how to control cloud costs went up, not down. There’s still much more maturing needed within enterprises about how to manage the cost of cloud more effectively. We also expected to see more enterprises take a cue from the clouds and start getting their own IT financial management skills polished up. We still have a ways to go here.
5. I&O will free the development teams to build apps in the cloud.
Grade: D. While we definitely saw some evidence of this in 2013, it was certainly not the rule in most enterprises. IT Ops in general continues to view the public cloud as an enemy rather than a welcome new member of the portfolio. As stated a year ago, developers don’t really need IT’s permission, and our surveys and client inquiries show the business-unit-aligned developers certainly aren’t waiting for it. Their use of cloud only went up in the last 12 months. While there are pockets of "type A" companies, generally those most aggressively embracing open source have IT Ops fully on board; as a rule I&O is still trying to understand what its new role will be in the cloud era.
6. We’ll get real about using the cloud for backup and disaster recovery.
Grade: B-. According to our 2013 Forrsights Hardware survey, DRaaS adoption is up to 23% already implemented, and DIY DR in the cloud is at 21% implemented. And we saw a strong rise in DR piloting, activities that are likely to become implementations in the coming year. We hope to see this jump in 2014, because our surveys also tell us BCDR is a key reason companies are moving to cloud — so where are all the success stories? BCDR analyst Rachel Dines: “They are coming — slowly — because DR is something people invest in in waves. Also, some may never be open to discussing it publicly because they see DR as a competitive differentiator.”
7. We’ll stop equating cloud with commodity.
Grade: A. Thanks to new cloud resource types containing SSDs, Infiniband, 10GbE, and even GPUs, the view that you could only get base commodity in the cloud simply went away. Look for even more proliferation in 2014. We are even seeing now specialized cloud platforms that go beyond commodity to deliver bare metal servers and HPC-specific tailored environments like the new GoGrid. In fact, you could argue the leading vendors are moving away from commoditization. AWS, IBM, HP, Rackspace, Google, and Microsoft … none of them plan to race to the bottom, and each is moving up the value chain with high-performance, specialized offerings.
8. We’ll stop equating cloud with AWS.
Grade: A. While AWS continues to be the dominant player in the public cloud platform space, 2013 saw significant competitive growth and a growing stable of legitimate private clouds. Microsoft Windows Azure added IaaS capabilities, which lit a turbo charger under its growth and adoption. Google, IBM, HP, and Salesforce.com ratcheted up their cloud platform plays significantly and saw strong growth as a result. And several leading enterprises showed they can deliver true private cloud value within their walls, such as Comcast, Concur, PayPal, CERN, MIT, and CTrip, who all spoke of their OpenStack-based clouds in Hong Kong.
9. We’ll acknowledge that advanced virtualization is a good thing, and no, it’s not a cloud.
Grade: C. The cloudwashing award will once again go to enterprise I&O departments who continued to relabel their VMware environments as private clouds. As our Forrsights 2013 Hardware Survey showed, European and Asian enterprises are the worst offenders, viewing cloud as just a tool for advancing their own virtualization agenda. However, this grade isn’t an F because there’s a bright side found in our Forrester Wave of private cloud software solutions. Through the 30+ customer interviews conducted for this research, Lauren Nelson found a variety of IT Ops teams who clearly understand the difference and are executing with this understanding. However, few of these environments offered self-service to the developer, fully-automated provisioning, standardized services, or cost transparency. Not because they don’t understand these requirements but because culture and organizational change is hard and takes time.
10. Developers will awaken to: Development isn’t all that different in the cloud.
Grade: B. An expanded cloud developer survey showed that while cloud developers are younger, more open source-forward, and more mobile-centric, they continued to use the same languages, frameworks, and development methodologies we use in the enterprise. There is much stronger adoption of modern application architectures and continuous delivery methods among the cloud developers and significantly higher adoption of Opscode Chef for workload configuration automation.
So what do we predict for 2014? Check back here next month to find out. Or subscribe to this blog now.
Forrester analysts Lauren E. Nelson, Dave Bartoletti, and Michael Facemire contributed to this report.