4 Steps to a People-First Strategy for Optimal DevOps
February 03, 2017

John Gentry
Virtual Instruments

IT teams are often tasked with managing the impossible: supporting the current needs of their organizations while continually developing applications and systems that reliably move business forward — oh, and let's not forget squeezing as much out of their budgets as they can, year after year with only nominal increases. Add to this the rapid pace of technology development and ever-increasing complexity caused by virtualization, hyperconvergence, BYOD and software-defined trends, and you get a perspective of the impossibilities that today's modern IT teams face.

This "new normal" can be thrilling, frightening or both, depending on your perspective. How enterprise IT manages compounding complexities, now and forever forward, affects overall business agility and innovation. With this inconvenient truth in mind, it's no wonder that DevOps and continuous delivery models are a primary focus. The performance and reliability of IT infrastructures delivering today's business-critical workloads is only as good as the accuracy and reliability of the systems monitoring and managing them.

While DevOps and continuous delivery models can (in theory) accelerate technology projects and help them run more smoothly and predictably, the required cultural realignment can prevent companies from reaping the full benefits. Developing and executing an effective strategy means shifting how teams collaborate and interoperate — and we all know how everyone loves change. Here are four steps that enterprise IT leaders are undertaking to facilitate the cultural adaptation and collaboration required to maintain IT and business agility:

1. Get buy-in, train, and enable data-informed collaboration

It seems so simple, but getting everyone on the same page must be the first, and sometimes the hardest, step. It is so important that it's worth repeating: Get. Everyone. On. The. Same. Page. The entire premise of DevOps and continuous delivery rests on the premise of bringing multiple teams together. Only then can you enable innovation, adaptation and transformation.

To make sure collaboration goes smoothly and delivers effective outcomes, focus on the basics. Start with development and operational functions, job descriptions, and similarities and differences. Determine the training and technologies needed to help with the cultural shift. Deploying analytics solutions that enable cross-domain collaboration is critical for aligning teams and helping them perform. These solutions help employees see DevOps transformation as an opportunity for growth, fueled by intelligence that they can consume, act upon and learn from.

2. Start simply, monitor comprehensively and decide confidently

There are several factors to consider when identifying which projects are good candidates to feed into the DevOps process, including overall scope, estimated costs and projected ROI. Start with projects that have a clearly defined and reasonable scope, as well as existing benchmarks. Determine success metrics for expected improvements, enable and monitor data-driven collaboration, and meet your targets.

This effort requires tight collaboration throughout development and operational organizations. Every team has relevant insights, which help form an accurate view of what it takes to deliver a successful project. It's crucial to have the right monitoring solution in place, one that provides a complete understanding of workload interdependencies, effects and how to optimize for overall performance.

3. Enable authoritative understanding and confident collaboration across silos

End users demand flawless experiences, and the ultimate charter for any DevOps team is to meet and exceed end-user demands. While DevOps teams must understand how new feature and functionality releases affect end users, they must also understand how these same releases affect the underlying IT infrastructures delivering them.

Effective DevOps models require close collaboration across multiple silos. To be effective, every collaborator must have an authoritative understanding of what's happening, and the confidence to act. End-to-end performance monitoring with algorithmically driven decision support provides teams with this accuracy and confidence.

4. Remember, DevOps has no "end state"

It's important to continually assess the training and cultural divides you are bridging. DevOps is a new way of life and it's just beginning. Everything is dynamic. Applications aren't static, and neither is the infrastructure that supports and delivers them. There must forever be a vigilant focus on understanding how organically evolving workloads affect the infrastructure. After new releases are put into production, everything must be constantly monitored and tuned for optimal performance.

Without an accurate, real-time understanding of end-to-end performance, every release cycle presents risks that can harm the DevOps process. It's critical to constantly gather and analyze performance data in order to determine how continually evolving workloads affect IT infrastructure performance. DevOps teams then need to use this data to proactively tune and manage infrastructures before emergent issues become actual performance problems.

John Gentry is CEO of Virtual Instruments.

The Latest

November 16, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 7, the final installment, covers IT Operations tools ...

November 15, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 6 covers DevOps and development tools ...

November 14, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 4 covers communication ...

November 13, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 4 covers more about combining Dev and Ops in teams ...

November 09, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 3 covers how to set up teams ...

November 08, 2017

Mainframes may be legacy, but they still run a lot of business. With so much in play, businesses run a great risk when practices for maintaining and developing on mainframe remain largely the same despite the rest of the organization undergoing significant change to keep pace with the latest DevOps trends ...

November 07, 2017

DEVOPSdigest asked experts from across the industry for their opinions on the best way to foster collaboration between Dev and Ops. Part 2 covers the personal interaction between Development and Operations ...

November 06, 2017

When you say "DevOps" one of the first words that comes to mind is "collaboration." But exactly how do we make this collaboration happen? This epic DEVOPSdigest list – to be posted in 7 parts over the next few weeks – intends to find the answer. DEVOPSdigest asked experts from across the industry – including consultants, analysts, organizations, users and the leading vendors – for their opinions on the best way to foster collaboration between Dev and Ops. Part 1 offers recommendations on how to get started with collaboration ...

November 03, 2017

It's 3:47am. You and the rest of the Ops team have been summoned from your peaceful slumber to mitigate an application delivery outage. Your mind races as you switch to problem solving mode. It's time to start thinking about how to make this mitigation FUN! ...

November 01, 2017

Throughout the software development and delivery process, each team plays a pivotal role in ensuring that the end result is exactly what is needed. However, all too often the software testing team, who come in towards the end of the cycle, is labeled as a bottleneck that slows down the process. But it is always important to see both sides of an argument and understand the pressures put on the testing team ...

Share this