Confessions of a VP of AppDev: “Agile - For Us That’s Still Waterfall, But With Jira”

Agile is now the norm with more than 90 percent of organizations practicing that methodology. However, it’s incredibly easy for DevOps to revert back to what looks more like a waterfall approach, albeit with Jira.

Sprint cycles can quickly get out of control with dev/test spending too much time blocking and tackling, and projects can suffer from shifting delivery dates.

There's also a vital piece to getting the agile process humming - data. In fact, data is often the largest factor that's holding AppDev teams back from truly unlocking the benefits of agile.

Sprint cycles only move as fast as data

While an agile methodology brings the promise of greater velocity and transparency to development, there’s a catch. The speed at which teams can securely access the data they need for dev/test usually holds them back from reaching their potential.

Scrum teams, who are often working within compressed sprint cycles that require a myriad of non-production environments, invariably spin their wheels, waiting on timely, relevant and fresh data to be populated. So while teams need more environments, they’re often starved of them because of having to spin environments up with the right data at the right time.

Sacrificing quality for speed

Agile teams often see testing as a bottleneck as they look to deliver on their commitments. Teams are so focused on demonstrating a compelling burndown chart at the end of the sprint, testing is usually one of the first activities to be sacrificed.

Testing teams, whether they’re performing unit, integration or overall system testing, face a shrinking time window to get their job done, so every hour counts. It’s also easy for administrative efforts to eat into time set aside for testing because sprint cycles are consumed by environment management, which is typically 25 percent or more of the cycle.

As a result, testing gets pushed to the end of the initiative, and subsequently gets squeezed out almost entirely in the delivery process. That’s why cutting environment set-up time to minutes is critical to preserving testing integrity during development.

Waterfall-like project slippage

Waterfall has always been notorious for endlessly slipping delivery schedules, in some cases stretching into years. Agile is an improvement with more visibility and accountability, but it’s not immune to slippage.

Without infrastructure to quickly spin up consistent environments that keep development moving, teams will end up with continuous rolling delivery schedules - ultimately impacting the ability to scale and meet business needs.

VPs of AppDev are looking to change up their projects by spinning up non-production environments with the latest data quickly and use data virtualization technology to cut data redundancy. It's a strategy that is enabling teams to speed up performance by 50 percent while also shifting left for testing.

Take Alberta Blue Cross as an example. Alberta Blue Cross, a Canadian-based company serving more than 1.7 million people with health coverage, wanted to modernize its IT systems to keep up with business needs.

But they faced several data-related challenges: 1) sprawl of physical databases rapidly driving up server, storage, software licensing and administrative costs, 2) database refreshes took one to two days to complete and 3) a new data model, business rule changes and a high volume of existing data meant that the performance and quality of data conversion scripts were critical.

By virtualizing their databases, Alberta Blue Cross was able to quickly execute a refresh of the complete environment. Developers worked during the day with the full environment refreshed every night through automated scripts. At the height of the project, the company had created nearly 18 environments containing more than 70 virtualized databases.

On-demand access to data coupled with fast refreshes allowed teams to reach implementation dates with higher levels of confidence. In addition to that, the environments’ built-in support of cutover and break-fix processes ensured a smooth transition to the new system.

What’s the takeaway? Growing data volumes, multiple environments and compressed sprint cycles mean that administrative slack has to be trimmed to maintain delivery and quality. Dev/test teams need the ability to populate their environments at speed with full production copies or subsets, all securely masked, so they can focus on what matters: delivering high quality software, on time.

Read about how to liberate your data and take a data-driven approach to testing with the Delphix Data Platform.