Choosing Infrastructure Extensibility to Accelerate DevOps Success


DevOps continues to generate interest as businesses and IT organizations seek more efficient ways of operating their data centers and enabling their developers. Organizations are realizing that competing effectively in today’s economy often requires delivering impactful features and capabilities to the business at an increasingly rapid pace. Many have looked to the principles and tools of DevOps to help them find these efficiencies, be more responsive to changing needs and seize competitive opportunities.

The low-hanging fruit in any organization’s DevOps strategy generally includes implementing more collaborative meetings between development and operations teams, regular assessments of short-term sprints and implementing new tools such as desired-state configuration platforms like Puppet, Chef, Ansible and Salt. These efforts generally include increasing self-service capabilities within their on-premises virtual environments through tools from VMware, OpenStack and CloudStack. All of these efforts tend to yield great improvements over previous operating models.

Unfortunately even this work is limited by the constraints of existing infrastructure, insufficient for delivering value in the new operating model. An evaluation of traditional infrastructure architecture will show frameworks with limited extensibility that force teams to make trade-offs and limit service offerings. These infrastructure components tend to offer a GUI and some type of application command-line interface (CLI) which might be leveraged for some amount of automation. This work is generally complicated, requires specialized knowledge of the CLI and working knowledge of writing expect or bash scripts. The biggest hindrance is that these resources are not easily shared or accessible to those not in the management group for that technology.

Rise of Infrastructure APIs

Organizations must now evaluate their infrastructure investments under a new lens, especially as they look at technology refreshes. Fortunately a wealth of new options, which provide feature-rich APIs built to communicate on known standards such as REST, have hit the market. The APIs for infrastructure enable a wide range of capabilities for integration beyond standard management constructs. For example, APIs allow developers and tool creators to leverage the capabilities of the infrastructure by making direct calls. These calls may allow changes to performance, network configurations, duplicating certain components, or managing the infrastructure end to end. This provides organizations the opportunity to work all the way to the hardware layer, thus reducing time to realizing necessary changes, communication between teams, and risk of misconfigurations due to human error or poor communication. Finally, it also allows organizations to respond quickly and programmatically as needs arise with more capabilities and less overall effort.

Integration with DevOps Tools

DevOps may be a cultural shift, but to be able to properly execute on the needs of the organization and the business, DevOps requires the right tools. Infrastructure APIs simplify the integration of infrastructure components such as storage, compute and network with the previously mentioned configuration management tools like Puppet. Using tools like Splunk, Loggly and Logstash with extensible infrastructure, users can more easily extract actionable data from the infrastructure. This data further simplifies the ability for teams to extract more value from their infrastructure with less implementation and day-two management efforts required.

Policy-Based Management

Organizations which also leverage policy-based management can take these gains to a whole other level as infrastructure APIs lend themselves to be easily managed via policies. Policies describe the constructs of the resources being presented to application servers, containers, hypervisors, or cloud management platforms. Extensible infrastructures then provide opportunity to create and present broader capabilities through policy-based management. This means developers can easily leverage these policies to once again minimize the communication required between the development and operations teams while also maintaining standardized offerings. This, in turn, improves operational efficiencies even more.

DevOps methodologies are enabling many businesses to operate with greater agility in order to compete in their respective industries. But the tools and cultural changes, while extremely valuable, are still limited to the underlying infrastructure. Traditional infrastructures are simply not up to the task and leave organizations wanting more from their technology. IT infrastructure now has the ability to help drive further efficiencies and capabilities and should be considered at each new refresh cycle. Utilizing infrastructure with robust APIs, integrations with common DevOps toolsets and policy-based management clearly allows DevOps-driven companies to accelerate their success. What can your infrastructure do for you?


About Author

Developer Advocate, NetApp SolidFire