Staff members may not be ready, either. For example, DBAs need to get comfortable with new ways of doing things, Chart said. “DBAs may find they need to insert themselves into the development process to help application developers who don’t know what they don’t know.” As a result, there is a lot to data governance that must be thought about early and tested thoroughly, he said. “When this is an afterthought, it can add significant risk to the business,” he cautioned.
The move to agile data organizations may also mean more moving parts as well, which is difficult for companies with already complex systems. “For the most part, these approaches actually make the delivery of data analytics more complex,” said Mike Maciag, CMO of Dynatrace, who pointed out, “Legacy solutions are not set up to manage these environments.” Maciag urged greater automation and collaboration. “By surfacing one common set of data, organizations can tackle this complexity with improved cross-team collaboration,” he advised. “Working with a single source of truth as you process these massive amounts of data becomes essential and makes it easier for these key teams to work together.”
Security issues may also arise as enterprises move to a more loosely structured architecture. Many organizations are still struggling to make the transition to a DevOps environment and hybrid or multi-cloud infrastructure, but these moves can bring their own security concerns, said Ingo Fuchs, chief technologist for cloud and DevOps at NetApp. “Traditional security tools are often ill-equipped to work in cloud environments, leaving IT teams struggling to keep up. The addition of containers and microservices, especially when it comes to complex hybrid systems sharing data in real time across multiple locations, can be daunting. That’s why building a solid data fabric with the correct level of governance and security for all stakeholders is critical.”
Seeing Results
There are a number of innovations and benefits that potentially can occur on many levels as agile data methodologies and technologies are put into place. There will be far greater levels of operational automation, for one. “Deployment of applications and services will become more automated than before—as opposed to a traditional sequence-based waterfall model of software development,” said Raghavan. “DevOps will ensure a continuous release of product and service features that will lend themselves to automated deployments within containers and microservices.”
The rise of containers promises to significantly change the way data and applications are moved throughout organizations—and they will be quite movable. The lift and shift from the on-prem world to the cloud are eased by simply putting applications into different containers and installing them on the cloud with an underlying orchestrator, said Raghavan. “In addition, each of the applications can be broken into smaller microservices containers that are now hosted in a distributed computing environment. Furthermore, using containers and microservices makes it easy to do cross-platform development. Gone are the troubles associated with developing applications that are uniquely suited to one kind of environment configuration versus another.”