Coping with Cloud Challenges
Hardly a week goes by without coming across news around the increase in the number of organizations moving their workloads to public clouds. In the digital world, embracing cloud is a fundamental requirement for engaging customers and conducting business at scale. Due to a multitude of factors, including compliance mandates, risk tolerance and IT capabilities, an organization’s infrastructure is more likely to be an eclectic mix of public and private cloud, and traditional on-premises – in short, hybrid IT.
Most organizations understand the cost and agility benefits of cloud and the need to embrace a hybrid model that enables them to meet a variety of specific business requirements. What’s less clear, however, is how organizations must constantly shape and refine these models to drive a sustainable business advantage. In other words, embracing approaches and technology that helps them to constantly exploit the cloud as a means to deliver what it should really provide – a way of future-proofing business.
So against this backdrop, it’s important to consider some of the many challenges that stifle an organization’s ability to fully leverage cloud:
- Outside-in performance monitoring. It took IT decades to learn how to manage the end-to-end application performance when it owned and could touch all of the underlying infrastructure. IT now has to manage that performance from the outside in. And this new challenge is arising just as applications are increasingly being used by customers—making their performance more critical than ever, both to the brand and the bottom line.
- Control of variable workload costs. When digital workloads ran exclusively on-premises, costs were generally stable. HR costs were fixed, hardware was capitalized, and software licenses were renewed annually. With cloud, unexpected spikes in demand can lead to unexpected spikes in cost. Enterprises must figure out how to manage these costs to know when they’re business-appropriate—and how to avoid them when they’re not.
- Value-based portfolio management. With on-premises IT, vendor relationships tended to be stable—in part because vendors could sustain functional advantages over longer periods of time. If and when it made sense to unseat an incumbent, hands-on access to the associated hardware, software, and data made it relatively easy to do so. However, cloud vendors leapfrog each other every day. This puts pressure on an organization to more frequently realign its cloud portfolio. And that realignment is tough to do, because IT no longer has direct access to either the origin or destination environments.
- API complexity. As enterprises stitch together a growing number of ever-changing cloud components in increasingly numerous ways, the interfaces between those components demand additional attention. But managing APIs isn’t like managing applications or servers—especially since you can’t dictate how they’re coded or provisioned.
In other words, cloud doesn’t simply shift workloads to rentable, low-cost infrastructure. It ratchets up business value—while also ratcheting up difficulty in the bargain.
How Will IT Cope?
(Image Source: Shutterstock)
These challenges continue to perplex IT departments. Now, IT isn’t just faced with managing more, but how to effectively co-ordinate a myriad of resources into cohesive digital processes without adding unnecessary cost and risk. Consider for example a decision to use public cloud services for testing. It’s a common use case for public cloud, but for businesses operating in highly regulated industries, how do they effectively transfer on premise customer records without exposing personally identifiable information?
Hybrid cloud challenges like these are forcing IT to adapt in many ways. Three adaptations are proving to be particularly important:
- New tools. IT can’t govern the cloud the same way as it governed on-premises resources. It must acquire new technology to enable more optimal operation given the increasingly complex and hybrid environments that encompass both cloud and on-premises resources in a common manner. One cloud-application for example could consist of a multitude of components, including many microservices supporting specific business functionality (each developed using different coding languages and using their own discreet data stores). In such environments, traditional tools designed for one specific technology will significantly increase the operation cost burden and never scale to support cloud models.
- New processes. Multi-stage escalation, annual budget cycles and other traditional IT mainstays don’t translate well in the dynamic, externally-owned world of cloud. IT must reinvent these processes so it can more reliably deliver services to internal and external constituencies—even as it has less direct control over the mechanics of service delivery. For example, rigid change management reviews and enterprise architecture dictates may need to be relaxed for newer style cloud applications that need to be delivered faster, change frequently and are shorter-lived.
- New culture. When you own your infrastructure, hands-on technical heroism is a useful cultural attribute. But it doesn’t do you much good when your infrastructure is elsewhere. That’s why IT culture must evolve to promote from ‘fighting technology fires’ to constantly crafting cloud models for maximum business value. For example, rather than waiting for failures across technology silos, IT must collaborate to ‘design for failure’ – that is, using monitoring methods to better predict and anticipate inevitable cloud technology failures so as to enact improvements in cloud design to better contain and recover from failures.
Cloud offers the enterprise tremendous opportunity – and its adoption, in some capacity, is essentially inevitable. But no one should think cloud will make IT any easier. We instead need to prepare for the management challenges it will continue to present, so that businesses can fully reap the benefits from cloud’s potential.
By Aruna Ravichandran