Multi-cloud – What Does Good Look Like
Everyone seems to be doing something around cloud, and we often get questions from our customers wanting to know if their current state represents progress, or if they are behind their peers. So I thought I would jot down some of the observations that we’ve collected over the years about what a “good” cloud environment looks like – what the architecture looks like, what the application portfolio will look like, how the IT organization will operate, and what metrics should be monitored. By the way, you can read about a recent analysis that IDG did of CIOs’ priorities and successes in this analysis of our transformation workshops. Some of the data in this blog comes from organizations who participated in these workshops.
In general, successful organizations 1) have built an infrastructure that securely connects and manages multiple cloud environments (public and/or private) using on-premises solutions, 2) deploy the right workload to the right cloud, and 3) have created a single operational model across clouds. They also 4) have a way to track and report metrics that prove the value of their investments.
What Does a Good Hybrid Cloud Architecture Look Like?
A hybrid cloud environment using VMware Cloud Foundation, vRealize, and NSX can enable you to seamlessly extend private clouds to public clouds as well as securely connect and manage multiple clouds with on-premises solutions. This type of architecture (see Figure 1) provides a way to run, manage, connect, and secure apps across clouds and devices in a common operating environment. It gives IT organizations choice and flexibility in how to build, run, and manage multiple private and public clouds. You can build and run applications on, and migrate them across, multiple clouds, securely connecting all clouds and managing all workloads across networks. Advanced operations management capabilities provide a unified view of the health, performance, and capacity management of workloads across clouds plus policy-based governance. It enables cloud brokering and integrating DevOps practices across multiple clouds. As needs change, it can provide an exit strategy for moving applications and workloads from the public cloud back on-prem.
How Will Your Application Portfolio Change?
Speaking of an exit strategy, how can you minimize having to move an app? Another quality of a “good” multi-cloud environment is that applications reside on the most suitable infrastructure. The graphic in Figure 2 shows where our transformation customers expect changes to their app portfolio over the next 2-3 years. For most organizations today, very little of their portfolio is cloud-native. High performing organizations will get to around 20% within the next 2 years. Up to 10% will be in SaaS applications like CRM or collaboration and communication systems. There will be significant portions of the portfolio that will benefit from the agility and automation of infrastructure as a service, and can remain on-premises, but moved to a new more modern hardware platform, such as a private cloud or to converged and hyper-converged infrastructure. In 2020, “good” will look like up to 75% of the portfolio on new platforms.
Additionally, most organizations will find that they can retire 20% of their portfolio. For companies who have been through an M&A, this can be as much as 50%. Application retirement can help archive and access the data while retiring an application and eliminate its associated costs for maintenance, licensing, and support. And lastly, depending on the industry, there will be some apps for which there is no cost benefit of moving to a new platform. These apps will remain in legacy infrastructure.
There’s a lot of value in taking a methodical approach to determining which applications belong on which platform. Use functional, technical, and business attributes for cloud selection with a consistent, repeatable framework for evaluation. Determine if you should migrate, modernize, retire, or simply leave an app as is. When we perform this service for customers, we will recommend a destination for each application that includes a relative cloud fit index describing how easily the application may be migrated there. We have a library of common targets, and we can also easily tailor the characteristics upon which we do the evaluation to customer-specific destinations, such as a custom private cloud or other modern infrastructure.
What Is the Value of a Multi-cloud IT Operating Model?
Changing the operating model is probably the most difficult part of transformation. Customers often ask where they will see cost savings if they do cloud right. Our experience is that you can reduce IT operational costs by roughly a quarter. It’s worth noting that these operational savings are not just from greater hardware and software efficiencies, as you can see by the large blue and the green sections on the graph in Figure 3. Those account for a relatively small portion of operating expenses. The greater savings come from more effective use of personnel and by leveraging automation and better tools to manage technology more efficiently. The key thing to remember is that it means working more closely with the business. It means an IT operating model that’s services and software product-oriented, not technology- or project-oriented.
What Are the Metrics of a Successful Multi-cloud Transformation?
Besides app disposition and operating model cost savings, what are some of the other metrics that result from a successful multi-cloud transformation? Take a look at Figure 4. Before we talk about the numbers, note that top performers have a strategy and roadmap that were agreed to by the top executives and business leaders of the organization. Read more in this blog series.
Most metrics in Figure 4 are self-explanatory and you can read them at your leisure. But let’s point out a few key goals that you should be reaching for – time to provision infrastructure of less than one day and virtualization of 80-100% (depending on your industry and number of ‘retained’ legacy systems). Another key part of successful multi-cloud transformations is the use of network virtualization – up to 50%.
For applications, top performers are deploying new software in less than a week. They are also using PaaS and CaaS systems for up to 10% of their development activities. And they have 80% of apps protected with automated recovery.
The best performers also have a service catalog and an IT self-service portal in production.
In summary, the graphic in Figure 5 highlights a few key best practices that we’ve learned working with clients on IT transformation programs around the world. Remember that a multi-cloud transformation is a tops down change agenda – successful projects have senior stakeholder involvement (e.g., CIO and CTO). And typically have a business case. These projects need to execute in multiple, concurrent workstreams in order to quickly show progress. A Transformation Program Office, or TPO, should be established to coordinate workstreams, as well as measure and report KPIs to show progress. IT infrastructure organizations cannot drive this process with a “build it and they will come” mentality. App owners from the business must be involved from the beginning.
If you want to see how your organization compares with successful multi-cloud environments, reach out to your Dell EMC representative.
See Related Blogs in the Series: