Blogs
Operating in a multicloud environment is a reality for most organizations today. Whether it’s human resources recruiting candidates, manufacturing tracking shipments, or marketing enticing...
Operating in a multicloud environment is a reality for most organizations today. Whether it’s human resources recruiting candidates, manufacturing tracking shipments, or marketing enticing customers, business units are doing an end-run around their own IT departments to directly access services on the cloud.
According to IDC, worldwide public cloud services grew 28.6% in 2017, with Asia Pacific the highest growing region in the world. IDG Research data from August 2018 shows that with 30% of IT budgets allocated to cloud spending, 43% of enterprises are using hybrid cloud, 12% are using multi-cloud and 30% are using both, with a complex evolution of cloud emerging as companies gradually move more applications and services to the cloud with an ever sophisticated and regulated environment.
Instead of ignoring or attempting to stifle organizational forays into multiple clouds, IT needs to facilitate, orchestrate and optimize enterprise multicloud footing. Here, I’m outlining what sets apart the CIOs that have mastered the many demands being placed on hybrid cloud environments by the enterprise.
By now most CIOs have had significant experience with Cloud and they’ve had some hard-earned lessons along the way. CIOs should now be familiar with the latest Born-on-the-Cloud techniques and technologies. They will have mastered some of the soft skills that allows them to manage teams, vendors and multiple internal stakeholders to drive the maximum impact from cloud spend without compromising security and/or compliance. Even the largest and most regulated banks in the world are moving to cloud. Earlier in 2018, Westpac, Australia’s oldest bank, was one of the first major banks globally that implemented a fully integrated hybrid cloud, moving the full spectrum of its banking services to cloud. The Bank has reduced the time it takes to set-up new applications from 19 days to just 3-5 days, with a more than 30 percent reduction in set-up costs from start to deployment of an application.
CIOs that have Mastered Cloud, have encountered and now predict the following challenges that emerged during the first phase of the enterprise journey to cloud:
- Predicting and Controlling “Hidden Costs” of Cloud. The CIOs that are mastering cloud are savvy to the previously “hidden costs” of cloud and they know how to work with vendors to accurately cost out cloud without falling into the trap of overlooking ingress and/or egress fees, the need to turn off un-needed infrastructure and overspend due to over-consumption. In response, we’re seeing CIOs being a lot more vigilant around transparency of these hidden costs and risks. This in turn is changing operating models, controls and commercials.
- Detecting and Preventing Shadow IT. With the democratization of cloud and SaaS, the enterprise has seen incidences of non-integrated, non-compliant, non-secure ‘Shadow IT’ enter the enterprise from the business units – from marketing software SaaS offerings, to HR apps and new collaboration tools. CIOs have had to find a balance between excitement, agility and value in new technology, and controlling and protecting the enterprise from security and compliance risks.
- Understanding the public cloud – dealing with Noisy Neighbours, Network Latency, Loss of Perimeter Security. CIOs with cloud experience have also seen the disruption that a “noisy” (aka CPU-hog) neighbouring application or cloud can have within the enterprise IT environment. CIOs that master cloud are now experienced at application architectures that are horizontally scalable and can have additional instances initiated to solve for ’noise’ issues to ensure every user of every application, critical or not, does not suffer from poorly behaved neighbors.
- Neighbourhood Watch: Incompatible neighbours. Some neighbours just don’t get on. CIOs that are experienced with managing multi-cloud environments have overseen enough run-ins to know that some cloud providers don’t make it easy to run an application suite across multiple clouds. Typically, security models, IP addressing schemes and connectivity need to be carefully solved for and generally call for implementation of a generalised integration plane that permits applications running on any one cloud to integrate with enterprise services – e.g. Identity and Access, Backup and Restoration, Security and enterprise Logging, Integrated bill-back or show-back.
- Avoiding Over-provisioning. Too much cloud? There is such a thing. CIOs are increasingly using tools that monitor under and over-provisioning to ensure that performance is not suffering nor is the business paying for unused capacity.
- Poor housekeeping. There are several traditional, auditable, steps in provisioning and decommissioning infrastructure – particularly in regulated industries. Audit trails of server hardening, recording of software versions and patch levels, association of applications running on servers, the availability requirements for those servers, registering them for Backup and Restoration and planning for disaster recovery. Keeping configuration databases up to-date is difficult in a non-automated world. In the world of cloud, these activities, the processes that support them and the storage of them must be changed to be completely automated as current manual processes will prevent any agility realised from turning infrastructure into code.
3 Emerging technologies to Help CIOs better Manage Multi-Cloud environments
- Brokerage – Brokerage, including the ability to instantiate an application and all its associated infrastructure and tooling, is an imperative to alleviating all the manual housekeeping tasks outlined above. The associated policies that ensure conformance with geo-fencing, proximity to data etc. can be executed by a work-flow component in the broker or be executed outside in an ITSM tool like Service Now.
- Provisioning – The provisioning capability for an enterprise must be robust enough to be able to materialize a hardened, secure, instance on anyone of a number of target platforms – on and off premise, Amazon, Azure, IBM Cloud, Alicloud or OpenStack. This permits the enterprise to optimize workload placement policies to maximize value while ensuring compliance and appropriate performance.
- Cloud Aware development methodologies – The recognition that traditional build and run processes, skills, KPIs and organizational constructs are ill suited to a multi cloud strategy is often left to the end. Building agile, interdisciplinary teams that include infrastructure coders along with application coders, QA professionals, security and business representatives is imperative to realizing the agility that companies enjoy in their Born-on-the-Cloud environments. Furthermore, the run operations are materially different – particularly for newly developed Cloud Native applications. The model is one of cattle, not pets – i.e. you recover from failure by restarting the application rather than working tirelessly to collect data, recreate the problem, issue a fix, run regression tests and push into production.
As the journey to the cloud continues for Government and enterprise, so too do the options, and the skills needed to optimise this new computing model for maximum effect. When managed correctly, cloud brings the promise of agility, cost effectiveness and elasticity that CIOs and business have been craving. As more applications and services move to the cloud, CIOs will need to sharpen their ability to both manage and optimise cloud for their enterprise.
General Manager, Global Technology Services, IBM Hong Kong