Predictions 2019: How Companies Can Improve Software Delivery


DevOps is not something you buy; it's a process you adopt, such as agile software development. DevOps is about constant improvement and is looking to go from DevOps 1.0 to 2.0 and up.

DevOps is a software development methodology that combines software development and computer operations. DevOps aims to shorten the system development lifecycle while providing features, patches and updates frequently, in close alignment with the company's goals.

In this eWEEK An article from Data Point, industry leaders at Delphix, Electric Cloud, IBM, JFrog, Pivotal, XebiaLabs and Datical, identifies the key areas that businesses should focus on in 2019 to continue to improve the results of the supply of software. Two common themes emerge: organizations need to move faster and do so by eliminating manual efforts. Automation, DevOps and agile development can all respond in different ways.

Andreas Prins, Vice President of Product Development, XebiaLabs:
"In 2019, developers will begin to break free of tedious, time-consuming tasks such as script pipelines by connecting their operations and CI pipelines, such as Jenkins, to the rest of the software delivery team and the company. DevOps pipeline. This integration will allow other team members to autonomously monitor the status of feature delivery and help developers reduce downtime to allow them to spend more time creating value.

Sunil Mavadia, Global Customer Care Manager, Electric Cloud:
"As organizations continue to rely heavily on DevOps in 2019, it's important to help them securely adapt to ongoing technology changes to stay competitive. To this end, we hear about five "big items" on their radar:

1. Cloud migration
2. automation
3. Artificial Intelligence / Machine Learning
4. DevSecOps
5. Upskilling

Sanjeev Sharma, Vice President, Global Practice Leader for Data Transformation, Delphix:
"As companies rapidly develop innovation by adopting DevOps on a large scale, they can now meet two major challenges:

1. Security and compliance: "Companies realize that it is important to continually offer smaller series of changes and validate security and compliance with each Sprint, rather than as a separate step just before publication. security teams as first-class members of their development teams, validation continues directly in their dev sprints.

2 Data Driven Applications: "DevOps practices are now embraced by data producers and consumers, enabling them to treat data as a code-like deployable asset, maximizing the ROI of their data-driven applications." data are better synchronized with code delivery cycles However, this field is still in its infancy Organizations will need to make the necessary changes in tools, practices and skills to manage, modify and collaborate around data , such as the code. "

Merker Kit, Vice President of Business Development, JFrog:
"The biggest challenges that DevOps will help solve are those created by the success of DevOps. More frequent updates mean more binary, storage and data files to manage – contributing to 44 zettabytes around the world by 2020. As we continue to generate more metadata about these files, about their origin , their behavior and their security, we can feed the machine learning Even more automated improvements and speed for DevOps. It may take a while before the promise of artificial intelligence in DevOps is fully realized, but we believe we will begin to see some interesting breakthroughs in 2019. "

Eric Minik, Product Management Manager, DevOps, IBM:
"The establishment of continuous distribution pipelines is no longer a marginal activity. This is pretty common and in 2019 we will see more attempts to get it through the business. I expect that this effort to adapt continuous diffusion (CD) will result in three major trends. First, approved enterprise toolchain selections that can drive CDs for each application. Secondly, and vice versa, other companies will adopt a multitude of CD solutions, standardizing a solution by large platform. One set of tools for the containers, another for the central computer, another for the Java applications and another for the databases. Coordination across this diversity will lead to the third major trend: increased emphasis on versioning and measuring the effectiveness and flow of diffusion in the enterprise.

"With so much data flow and the emphasis on versioning decision making, look for more artificial intelligence capabilities in your tool chains."

Dormain Drewitz, Senior Director, Product Marking, Pivotal:
"By 2019, more and more traditional DevOps adopters will benefit from automated patches as patch cycles shrink from several months to several weeks." Software-defined networks will allow more businesses to create and maintain patches. to exploit application platforms whose network layer is treated as code Although developers can not totally ignore the network (latency and unreliability of the network are realities), the Network as a code will help DevOps teams tremendously, and we will finally see scattered examples of cloud data architectures from non-Internet companies emerging.The domain-based design will provide a common foundation for those who advance DevOps' practices of including data. "

Robert Reeves, co-founder and CTO, Datical:
"From my point of view, we will definitely stop hearing about" DevOps. "Or, I hope so! As the adoption of DevOps will increase, we will see the adoption by teams that we had not even thought of during the first iteration of DevOps.This follows the same path that agile when we started to see things such as IaC (infrastructure as a code) and will be guided by the obvious advantages that these teams (security, database, network) see in other areas, and we will see another security flaw directly attributed to manual changes and uncorrected dependent libraries. Sigh …

"Tom Petty was wrong. "Waiting" is not the hardest part. Continuing to improve is the hardest part. DevOps is not something you do once and claim victory. As if you were going to the gym and eating better, you must continue to search for and eliminate areas of manual stress. And, yes, running a script is a manual effort. From application to infrastructure, database and security, we are immersed in manual efforts that DevOps can remove. Simply because it is the correction of the database or system or security and was previously considered "too important to be left to the machines", we must change this by thinking that it is "too important to leave to humans".

If you have a suggestion for an eWEEK Data Point article, send an e-mail to