Today’s data center architectures are becoming more complex and increasingly software-defined, making them increasingly difficult to manage without a strong automation strategy.

Agile processes have brought great changes to the supporting application development landscape. Now, we must manage a complete, end-to-end pipeline starting with new coding and faster practices, and then move into new testing and QA experiences. This then leads to expanded infrastructure platforms like containers, and finally to new and constantly changing devices for the end-user experience.

As if the changes in the development and deployment of applications weren’t already challenging enough, IT is now being asked to support end-to-end software-defined data centers, from network to containers to hybrid cloud server farms. This adds new capabilities at the expense of additional complexity, and it’s this complexity that is driving the importance of dedicated automation specialists and creating new opportunities for DevOps.

All of these new opportunities are placing an unprecedented amount of stress on limited IT staff and infrastructure resources. To help address these new and ever-changing requirements, IT professionals have been addressing the problem one issue and one tool at a time, often not thinking about an overall automation strategy.

Tools like IBM Bluemix, Jenkins, Maven, HashiCorp, Chef and Puppet help manage the new, complex application life cycle, but are standalone solutions that don’t offer a lot of synergy.

In developing an automation strategy, one of the first questions to ask is, “What does automation mean to me and my organization?” Even Gartner has many different categories for their automation:

With all of this, where should you start? It’s important to break the problem down into smaller pieces while keeping the overall automation strategy in mind. Start from either a bottom-up or a top-down approach. A bottom-up approach might look like:

  • Network vendor tools to manage the software-defined network and storage
  • A hybrid cloud management tool
  • Runbook and IT process automation
  • Job and batch scheduling
  • Service catalogs and end-user self-help
  • Performance automation
  • Configuration management tools
  • Developer testing and QA tools

Try to use tools and pick vendors that have robust API integrations, and are committed to open standards and open data stores of their data. Unfortunately, there isn’t a single vendor that can solve all the automation needs across IT, so we are having to pick and choose, and hope that the vendor has supporting integrations or we have the skills to do it in-house.

One of the outcomes of all this is the advent of the DevOps specialist. As the picture below shows, IT has traditionally had the Business Analyst and Operations Analyst roles in various forms; what is new is the defined role of DevOps. This role exists in some sort of “tools” group and should have the lead in establishing your automation strategy.

 

This gives IT the resources to create a “tools” team led by the DevOps staff, which can develop a comprehensive and cohesive automation strategy.

Automation, like any other functional area in your data center, needs dedicated professionals, a clear mandate from management, and strong leadership to ensure success in this important but often-overlooked area within your data center.

Sirius IT consultants can help you deliver on the promise of automation across your data center, from Agile and QA teams to tools like AppDynamics for measuring application performance, to the creation of service catalogs in your favorite service deck like ServiceNow, and even managing complex configuration management solutions in Chef or Puppet.

Contact us today to learn the value Sirius can bring to your IT organization.