Shift focus from CI Pipelines to DevOps Assembly Lines
Pipelines shows a complete visualization of the workflow from the version control system to the production and serves as only pure CI ( continuous integration ).
It is much more convenient and productive when CI and CD ( continuous delivery ) in one bottle (as in Assembly Lines ). Why? Yes, just a software company saves time and effort spent on the process of developing software, and controls all stages of work. In 2019, a full transition from CI Pipelines to Assembly Lines is expected .
DevOps tools in AI and ML
Web forms are the weak link in any automation. Everyone is afraid of manual forms, but organizations use them when they want to get data from customers, employees, or partners.
Machine learning technologies, as well as AI, are already capable enough to reliably process voice, video, text or image. This means that forms can now be filled out with a phone call, videotape, photo, etc.
Of course, the error rate with such a filling will be higher, but it will depend on the level of training of the AI model and how well real data is used to improve it.
Automation will be the main focus
DevOps talks a lot about automation. The ideal option is a zero-touch process (but this is still in perspective). Understanding DevOps-cycle ” 6C ” and software automation cycle between these stages is the key, and it will be the main objective in 2019.
From 2019, data usage will be closely monitored, as the GDPR (General Data Protection Regulation) becomes a reality, and similar privacy laws are being adopted around the world. To implement such control will require labeling information. Therefore, tools related to managing and tagging data will be crucial in tracking their origin and use.
Confidential information must be encrypted. Hence the responsibility for managing encryption keys (they must be rotated and stored safely). Services that provide such functionality will be widely used in the world of DevOps.
Testers will learn to code
Testers who can program and write scripts for testing are in great demand in DevOps. If you are a tester and do not know whether you need to study programming, we recommend that you do it. Understanding various DevOps tools and automation scripts today plays a vital role in software development.
Manual testing will become obsolete in 2019, because it requires an unreasonable amount of time. Automation in testing not only increases efficiency, but also ensures that the product gets to the market faster.
Container Development and Data Location
Enterprises working with SaaS or PaaS use public and private cloud storage in different countries. As a result, important data and applications will be deployed in different geographic regions, which will complicate the deployment process.
Development on the basis of containers will become commonplace in the enterprise. This approach has many advantages: creating a distributed workload, increasing the demand for sophisticated monitoring tools and technologies, etc.
The growing popularity of microservice architecture
DevOps and microservices go hand in hand. Microservices are independent entities and, therefore, do not create dependencies and do not violate child systems when something goes wrong. Microservice architecture helps companies easily deploy and add new features.
It is expected that companies will switch to microservice architecture to speed up the development and production process.
Compiler – the new king of data centers
The death of Moore’s law does not help the processors to keep up with the increase in workloads. To bridge the performance gap, modern hardware, such as graphics processors, FPGAs and ASICs , must be systematically updated.
Developers will start using higher-level languages such as OpenCL and CUDA , and compilers will generate code to run on different hardware depending on the input of the deployment level, including workload, hardware configuration, and data size. DevOps practitioners will start “making friends” with compilers so that the code can be created just in time , as close as possible to deployment.
Kubernetes has become the fastest growing container technology due to its functionality and ease of use. This tool acquired a large opensource-community, and many CIOs and system architects began to use it in depth throughout the world.
In 2019, the rapid growth of development and expansion of this technology is expected.
CI / CD allows you to use fast changes daily to meet the needs and requirements of customers, so you need to take into account all the approaches to safe development at the software design stage.
Recently, there has been an increase in DevSecOps among the DevOps trends. This is the introduction of “security” into the application development life cycle, to reduce the risk of occurrence and increase the likelihood of detecting vulnerabilities. The model assumes that each link in the development is responsible for security, which means that it is easier to identify who did what and what went wrong.