DevOps automation for legacy and physical infrastructure

Posted by admin August 4, 2015
DevOps automation for legacy and physical infrastructure

Starting a business in a garage, dorm room or tiny upstairs office complex provides very few actual advantages, despite the idealized picture of this approach seen in the histories of tech firms such as Apple, HP and Facebook. However, there is at least one notable benefit to this bare-bones strategy: The company will likely have far less legacy IT infrastructure than if it had initially set up shop in a more traditional office setting.

For the very few startups that mature into successful Web-scale organizations, this freedom from old physical systems, no matter how it is attained, can provide a big leg up on the competition. For example, they can build their operations upon virtualized and cloud assets that are more easily adaptable via software updates and procurement and maintenance of commodity hardware. This strategy lends itself to outstanding business agility, while providing the technical foundation for organization-wide movements such as DevOps that can help reinforce its advantages.

DevOps automation and cloud orchestration are the bread and butter of the modern agile IT organization. But conversations about these topics are often overwhelmingly biased toward the newer aspects of IT, namely how to work with virtual and cloud infrastructure that is naturally suited to the rapid releases, cross-organizational collaboration and continuous refinement emphasized by DevOps. What about the legacy and physical assets still powering so many mission-critical applications at so many organizations today?

The DevOps difference: More frequent deployments and more reliable software
Puppet Labs recently released its annual State of DevOps report, which provided some insight into how DevOps can benefit organizations under ideal conditions. The report's preparers noted that DevOps practices can "improve both throughput and stability, leading to higher organizational performance."

They also devoted attention to how DevOps should ideally apply not just to new applications but to existing ones as well ("greenfield, brownfield or legacy," in their words). As long as testability and deployability are key metrics for measuring a given program's efficiency, DevOps can play a role in potentially improving its performance.

Cloud helps enable DevOps for old and new applications alike.Cloud helps enable DevOps for old and new applications alike.

This is good news for organizations that are not in the enviable positions of the hyperscale giants that have always been free of legacy infrastructure. Moreover, DevOps has sometimes been portrayed as being out of reach for IT teams accustomed to siloed departments and waterfall methodologies, but with the right tools and cultural adjustments it can benefit them, too.

"Using business automation for service orchestration, businesses can automate the delivery of simple or complex IT services requested by their users," explained Ben Rossi in an article for Information Age. "By orchestrating the provisioning and changing of different service components across business and system layers, they can evolve existing legacy data center assets towards innovative cloud operations - without ripping and replacing existing IT investments and skills."

DevOps' role in bimodal IT
This balance between old and new that Rossi describes is akin to what research firm Gartner has termed "bimodal IT." Basically, this phenomenon involves maintaining legacy "Mode 1" applications even as the organization broadly shifts to newer "Mode 2" programs that take fuller advantage of virtualization and cloud computing.

"DevOps automation can support the modernization of Mode 1 applications."

DevOps automation enabled by a solution such as QualiSystems CloudShell can support the modernization of Mode 1 applications. In practice, this overhaul may involve the setup of an efficient private cloud with a self-service interface, with infrastructure-as-a-service helping in the transition away from the protracted waterfall cycles of the past:

  • All teams, including security, devtest and compliance, get access to production-like environments via a convenient portal-driven self-service interface.
  • Infrastructure environments can be obtained in minutes rather than days or weeks; private cloud IaaS essentially brings the on-demand resource allocation of the cloud to internal processes such as devtest.
  • Software outcomes become better over the long term thanks to more regular commits, shorter development cycles and more democratic, simplified access to on-demand resources by all parts of the organization.

DevOps has sort of a symbiotic relationship with bimodal IT. That is, its emphasis on collaboration and cloud helps supports the broad modernization of existing Mode 1 and upcoming Mode 2 applications. At the same time, the organization's focus on cloud infrastructure helps sustain DevOps across both virtual/cloud and legacy/physical assets, preventing DevOps itself from becoming another silo holding back agility.

The takeaway: DevOps can be applied to all types of infrastructure, not just newer virtualized and cloud-based assets. It may have an association with hyperscale organizations that have legacy-infrastructure environments and the ability to rapidly develop, test and deploy software, but it can also be implemented even in organizations with older processes and servers still in place. A platform such as QualiSystems CloudShell can provide the technical basis for implementing private cloud IaaS and DevOps to streamline the management of legacy, physical, virtual and cloud infrastructure.