By Pierre Moncassin
Several times over the last year, I have heard this observation: “It is all well and good to introduce new cloud management tools — but we need to change the IT roles to take advantage of these tools. This is our challenge.” As more and more of the clients I work with prepare their transition to a private cloud model, they increasingly acknowledge that traditional IT specialist roles need to evolve.
We do not want to lose the traditional skills — from networking to storage to operating systems — but we need to use them in a different way. Let me explain why this evolution is necessary and how it can be facilitated.
Emergence of Multi-Disciplinary Roles
In the traditional, pre-cloud IT world, specialists tended to carve a niche in their specific silos: they were operating systems specialists, network administrators, monitoring analysts, and so on. There was often little incentive to be concerned about competencies too far beyond one’s silo. After all, it was in-depth, vertical expertise that led to professional recognition — even more so when fast troubleshooting was involved (popularly known as “firefighting”). With a brilliant display of troubleshooting, the expert could become the hero of the day.
In the same silo model, business-level issues tended to be handled far away from the technologists. The technology specialists were rarely involved in such questions as billing for IT usage or defining service levels — an operations manager or service manager would worry about those things.
Whilst this silo model had its drawbacks, it still worked well enough in traditional, pre-cloud IT organizations — where IT services tended to be stable and changes were infrequent. But it does not work in a cloud environment, because the cloud approach requires end-to-end services — defined and delivered to the business.
Cloud consumers do not simply request network or storage services; they expect an end-to-end service across all the traditional silos. If an application does not respond, end users do not care whether the cause lies within networks or middleware: they expect a resolution of their service issue within target service levels.
Staffing the Cloud Center of Excellence
To design and manage such cloud-based services, the cloud center of excellence (COE) requires broader roles than the traditional silos. We need architects and analysts who can comprehend all aspects of a service end-to-end. They will have expertise in each traditional silo, but just as importantly, the ability to architect and manage services that span across each of those silos. I call these roles “cross-domain experts,” because they possess both the vertical (traditional silo) and horizontal (cross-silo) expertise, including a solid understanding of the business aspects of services.
Cross-domain competencies are essential to bring a cross-disciplinary perspective to cloud services. These experts bring a broad spectrum of skills and understand the ins and outs of cloud services across network, server, and storage — as well as a solid grasp of multiple automation tools. Beyond the technical aspects, they are also able focus on the business impact of the services.
Cross-domain experts also need to cross the bridge between the traditionally separate silos of “design” and “build.” Whilst in the traditional IT model the design/development activities could be largely separated from the build requirements, a service-for-the-cloud model needs to be designed with build considerations up front.
Every team member in the COE needs to possess an interdisciplinary quality. If we look more specifically at the organization model defined in the white paper Organizing for the Cloud, after the leaders, these hybrid roles are foremost to be found in the following categories:
- In the tenant operations team, the key hybrid roles are service architect and service analyst.
- In the infrastructure operations team, the architect is a key hybrid role.
Takeaways
- To build a successful cloud COE, develop multi-disciplinary roles with broad skills across traditional silos (such as networks, servers, and middleware). Break down the traditional barriers between design and build.
- Foster both formal training and practical experience across domains.
- Organize training in both automation and management tools.
——-
Pierre Moncassin is an operations architect with VMware Operations Transformation Services and is based in France. Follow @VMwareCloudOps on Twitter for future updates, and join the conversation by using the #CloudOps and #SDDC hashtags on Twitter.
