Insights

Five ways to add agility to your IT infrastructure

Aditya Vadaganadam, Agile Coach at UST |

In my previous post on IT Infrastructure Agility, I explored whether a lack of enterprise-wide adoption of new ways of working was responsible for failed Digital Transformation initiatives. I discussed how DevOps, Agile, cloud, and lean startup thinking have introduced a tectonic shift in how enterprises work. For infrastructure leadership and management, this is nothing short of a perfect storm that they need to navigate.

Organisations can overcome the barriers to infrastructure agility, especially now when they can leverage proven tools, techniques, and new ways of working to add agility to their IT infrastructure.

I’ve listed five big themes of the enterprise digital transformation initiatives that I have worked on over the years.

Reimagining Your Legacy Technology Stack

Technologies such as the AS400 or mainframes that have been the bedrock of technology for decades have also been culprits for constraining agility.

Many organizations have realized that replacing legacy systems with open system stacks is not an easy feat and instead must innovate on what they have. IBM’s continued increase in mainframe sales over the last many years is a reminder that the legacy stack is sticky!

Naturally, IBM and other members of the old guard are driving innovative ways to help enterprises manage their legacy stack and to keep legacy relevant. Organisations like IBM,https://www.compuware.com/

BMC,Microfocus,Broadcom, and ARCAD bring modern engineering practices such as continuous integration/continuous delivery (CI/CD), automated code analysis, virtualization, lightweight integration, and test automation to these so-called legacy platforms.

These new practices not only give these systems a new lease on life, but they also enable modern enterprise systems to be built iteratively and incrementally—and not be handicapped by previous architectural constraints anymore. There is no legacy system, just legacy thinking!

Embracing Radical New Operating Models

Infrastructure agility requires new operating models. While there is a multitude of patterns out there, each organization will require a slightly different approach. A few of the most popular patterns are:

1. Transparent Shared service: Applying lean, value-stream mapping and Kanban techniques to emphasize transparency, business-aligned prioritization of work, and limiting work in progress to the existing shared service based constructs.
2. Responsive shared service: application squads “seconded” specific infrastructure members for more responsive service. The secondment model’s limitation is in the ability to scale beyond a few iconic programs of work.
3. Cross-skilling: Infrastructure and application team members learn skills across each other’s disciplines to make delivery teams self-reliant.
4. Disrupting with cloud: Disrupting an organization’s traditional infrastructure capabilities by aggressively pursuing private, public, and hybrid cloud capabilities.

Each pattern requires a different approach and a different set of criteria to make it worthwhile to the nuances of your organization. These are non-exclusionary, so organizations can use a combination of these approaches to be effective.

Investing In culture and people is a key common assumption for your radical new operating model to work!

Pursuing Lean, Agile, and DevOps puts a spotlight on infrastructure teams like never before. No longer operating in the shadows and behind a queue, they must be increasingly customer-focused, collaborative, and value-seeking. They are also held more responsible for driving the direction of the organization than ever before.

Managers need to empower infrastructure teams to make key decisions versus the traditional structure of management dictating what teams must execute.

Insourcing ‘Responsive’ Infrastructure Capabilities

Almost twice as many CEOs are intent on building up in-house technology and digital capabilities (57%) as those plan on outsourcing it (29%), according to the 2017 Gartner CEO survey.

The time has come for organizations to hire employees for in-house technical roles and build near-shore hybrid teams.

The rise of digital transformation prioritizes a direct connection of management with an organisation’s IT department, which is strengthened by having resources internally.

Inhouse IT departments are more likely to better handle the pace and flexibility required of organizations that have embraced digital transformation.

COTS 2.0 (Smarter and open commercial off the shelf products)

COTS products are an integral part of many enterprise DNA - be it the omnipresent SAP product suite, emerging products such as Salesforce, or seemingly arcane and heavily customized COTS products!

Consider Intelligent Enterprise Resource Planning (i-ERP) Software as an example.

In 2025, SAP/R3, the largest ERP product in use, will no longer be supported officially, and enterprise planning programs (albeit in a waterfall way) will move to SAP HANA Enterprise Cloud, the company’s new generation ERP platform.

Product vendors in the ERP space have embraced cloud and API-led integration—IDC calls this “intelligent ERP” or “i-ERP.”

Unlike traditional ERP, i-ERP will maintain pace with modern engineering practices and flexible architecture fundamental to infrastructure agility.

Everything as code and the API mandate

The idea of ‘everything as code’ extends application as code to all layers of the stack - even infrastructure components such as compute, network, operating systems - and at scale! The ability to provision, utilize, and tear down entire data centres consisting of thousands of servers using automated orchestration and provisioning capabilities is not a pipe dream but the reality of many early adopters. Even traditional enterprises have embraced the idea - consider Walmart, historically a conventional brick and mortar retailer, as of 2016, deployed and managed 55,000 servers using fully automated and orchestrated practices.

‘Everything as code,’ when supplemented with another emerging practice of the API mandate, enables true autonomy for teams.

What is the API mandate?

Jeff Bezos’ famous memo best articulates the API mandate in the following principles:

1. All teams will henceforth expose their functionality through service interfaces
2. It doesn’t matter what technology is used. HTTP, Corba, Pubsub, custom protocols — doesn’t matter
3. All service interfaces, without exception, must be designed from the ground up to be externalizable
4. Anyone who doesn’t do this will be fired

A few brave traditional enterprises and their infrastructure divisions combine ‘everything as code’ and the API mandate together to enable autonomy and de-coupled scaling.

Summary

These are some practical approaches to help achieve infrastructure agility - embrace new ways of working and radically different operating models, adopt modern engineering practices and technology, elevate your people, and optimise your legacy tech.

To continue this conversation about infrastructure agility, contact the author:

Aditya Vadaganadam

UST email - aditya.vadaganadam@ust.com

LinkedIn - https://uk.linkedin.com/in/adityavadaganadam