When you’re implementing a transformative technology-based solution, often times the easy part is the technology integration, and the hard part is people and process change. That can apply to cloud computing as well particularly if you’re a large enterprise. Large enterprises have IT operating models and processes that tend to calcify over the years, along with the organizational fiefdoms and siloed IT staff roles that evolve with it. Unfortunately, these and other limitations can make cloud adoption that much harder.
Challenge #1: Breaking down IT fiefdoms and silos.
It shouldn’t be a surprise to hear that IT organizational silos can be a hindrance, but you may be surprised how quickly you’ll encounter those hindrances. For example, setting up an initial private cloud will likely involve separate IT groups responsible for storage, computing, networking, platforms, and security. Coordinating between them is one issue, but what makes this even more challenging is that the cloud vendors today almost universally won’t have interoperability issues worked out between products. You’ll encounter documentation gaps, integration problems, incompatibilities, and you’ll be escalating issues through multiple vendors which can result in long delays. This can easily result in nasty finger pointing and bad blood between those IT fiefdoms that are under deadline pressure and not used to working this closely before.
Above is just one initial example. If you’re taking the long view of enterprise cloud adoption (and we hope you are), then you’re interested in changing your IT operating model to be more agile and demand-driven. That can include new continuous delivery models and implementation of fully governed and standardized “as-a-service” offerings across dev, test, IT ops, and others. All of which implies multiple, sustained challenges to established IT silos and their traditional boundaries. Be ready to start adapting early.
Challenge #2: Failure to adopt policy-driven governance from the beginning
Cloud initiatives are a new opportunity to drive major process automation improvements and implement new governance best practices. However, this window of opportunity will close if you don’t implement solutions for policy-driven governance and lifecycle management early enough.
Just because it’s technically possible to move workloads anywhere doesn’t mean it’s legally permissible to do so. Providing governance early is important, because user demand will be there even if the policies are not. You need to put sufficient safeguards in place before a governance failure creates a blame game between teams or jeopardizes the whole project. To do this, you’ll need a policy engine integrated into a cloud management platform that’s flexible enough to capture policies from industry regulations to user-specific consumption limits, and then enforce them in a granular way across organizational groups, projects, and environments.
Challenge #3: Groups that cling to legacy systems, old platforms, and proprietary technologies
Organizations may stick with older systems and platforms under the premise of “if it isn’t broke, don’t fix it.” Even though the application might function okay, the maintenance and support costs may be “broken” and exceedingly high, along with other problems like lack of scalability.
Many enterprises use cloud initiatives as a vehicle to drive standardization of as-a-service offerings to lower maintenance and support costs in addition to other consistency and automation benefits that come through standardization. Count on some organizations to vehemently resist the move toward standardization, with potential battles for legacy system exceptions and the push for standardization getting potentially ugly.
Challenge #4: Effects of short-cutting the enterprise vendor selection process
It’s risky, but very easy, for employees across the organizations to start using different cloud services immediately with a credit card. It’s much harder to select and enforce the use of enterprise-wide cloud vendors that can meet a cross section of enterprise needs. The latter requires different IT fiefdoms to roll-up their sleeves and work together on enterprise requirements and vendor due diligence. Although more challenging, the end result is better for the enterprise with greater management control, cost saving, risk mitigation, and visibility.
Reviewing cloud vendors takes a lot of work today. Skepticism is definitely encouraged, and you should expect a lot of hidden complexity and integration work underneath most cloud software vendor claims and UIs. Cloud service providers require similar due diligence, and you’ll need to look beyond pricing and SLAs to their cloud provider APIs. Not all provider APIs are created equal, and several enterprise use cases are impossible to support with some public cloud APIs. You’ll need to know which ones impact you.
You must be logged in to post a comment.