Five hot and four cooling trends in infrastructure and operations

The IT world is constantly changing, with new tools and strategies emerging to disrupt the status quo. At some point, technology pioneers will win the market competition. And many times, this change is more like a pendulum, swinging back to where it was.

Infrastructure and operations departments have also seen some changes, but at a slower pace than other areas of technology. Teams responsible for managing code and keeping systems running smoothly are naturally very cautious. Experimentation and change for the sake of change are inevitable choices for those who are innovators. When enterprises want their businesses to run smoothly, it is more important to maintain the stability of infrastructure and operations.

[[315313]]

However, recently, a number of new strategies and tools have emerged that have changed the heavy lifting that infrastructure departments do in keeping servers and networks running. Some of these trends are driven by innovative technologies, some by economics, and some by political realities. All reflect the demands on maintenance teams to provide greater security and speed without sacrificing stability.

Hot 1: Cloudy

The advantages of moving code out of on-premises data centers and into the cloud have long been recognized. Renting servers maintained by others is ideal for scaling computing resources and workloads. While there are always questions about trust and security, cloud computing vendors have addressed these issues through dedicated teams with economies of scale.

If cloud computing is a good idea, why not two or three or more? Supporting multiple clouds may be more work, but if your developers are careful when writing code, you can eliminate the danger of vendor lock-in. Your finance staff also has the opportunity to benchmark their software across multiple clouds to find the lowest-cost cloud provider for each workload.

Trend 1: Dynamic Websites

The global Internet was originally made up of static files. A web server received a URL and responded with the same file. This simple mechanism quickly became obsolete when developers realized that they could customize what users saw when they visited a specific URL. Users liked personalized pages, advertisers liked the flexibility of targeting, and businesses liked the opportunities that dynamic websites brought. Sophisticated frameworks could help create customized pages for anyone who wanted them.

This trend has changed again in recent years, as developers and businesses have realized that despite all the options, most web pages end up looking pretty much the same to everyone. Is the cost of creating smart server logic worth it? Why not use all the speed of an edge-savvy content delivery network to send the same data to everyone? Some of the latest web development tools can now pre-fetch your site into a folder of static web pages, giving you all the flexibility of a dynamic content management system with the speed of static files. However, the result is not completely static, as using AJAX calls you can use JavaScript to fill in some gaps or collect some custom data.

Hot 2: On-premises cloud platforms

As part of their sales pitch, cloud computing vendors have been pushing open data and code, asking users to trust them with their data, and while this approach does allow customers to place certain restrictions on where code is hosted geographically, users don't need to know what's going on on the computers they rent in the cloud.

However, some businesses want to store and process data themselves, which feels safer, and some businesses need to protect their data at a higher level than others. The solution? Running the cloud company's software and tools on its on-premises servers. Setting up the instance feels like a cloud platform, which combines the flexibility of a cloud virtual instance with the security of physical control over the server. In addition, if the company can control the additional costs of installing and maintaining the hardware, this approach can sometimes be less expensive.

Cooling 2: AI is everywhere

A few years ago, when AI applications were growing rapidly, many businesses rushed to adopt AI systems. As businesses adopted AI and collected data points, huge data sets emerged. More information means more training opportunities for AI, and should produce smarter, more accurate results.

This overreach has raised alarm bells. Many are beginning to realize the privacy threats posed by collecting the vast amounts of information needed to exploit AI. Others worry that the datasets being accumulated are imbalanced and biased, making it more likely that their AI will simply learn to respond to that bias. Others worry that AI could become too powerful, controlling large parts of the decision-making chain. Now, AI developers have to do more than just answer whether the job can be done. They have to weigh the various dangers and consider whether the job should be done at all. This has also led to a growing demand for "explainable AI."

Hot 3: Serverless

For a long time, developers have wanted to have full control over their environment. This is because if they can't specify the exact distribution and version, they can't guarantee that the code will work properly. Many people know this is a contradiction, so they want to have root access to the computer they control.

All the copies of the same files might keep everything running smoothly, but it's inefficient and wastes resources. Serverless tools take all the hard work out of the system. Now, developers only need to program a simple interface that will load their code when needed and then charge users. For jobs that run occasionally, whether it's background processing or a website that doesn't get much traffic, they don't need to take up a lot of server resources, which has a full copy of the operating system and doesn't take up any memory.

Cooling 3: Building your own components

Developers typically build software by combining a series of small components and libraries. Each part contributes something to the overall package. Many components are off-the-shelf products, such as databases or popular APIs. It is not uncommon for dozens or even hundreds of components to work together to provide a unified web presence to the user.

However, these products have been slowly getting smarter lately as developers add more functionality. For example, some databases are more tightly integrated with the network, and they offer the ability to sync data stored on the client, eliminating the need to build this functionality. Features like translations can now be integrated into other tools. As applications and services grow, the glue code and customization disappears, and sometimes it becomes a configuration file. Flowcharts still include the same functionality, but are now much more feature-rich.

Hot 4: Green and energy-saving artificial intelligence

Over the past few years, when it comes to machine learning and artificial intelligence, the more compute and the more training data, the better. If you want to get the most out of artificial intelligence, then scaling up is the way to get better results.

However, computing requires a lot of electricity, and many businesses are beginning to question the power-hungry algorithms, prompting AI developers to test whether they can save more power without significantly increasing electricity costs.

Cooling 4: Basic Repository

In the past, a code repository didn’t have to do much to keep up. If it kept a copy of the software and saved information about changes over time, its storage would quickly grow. Today, developers expect repositories to push their code through a pipeline that might include everything from basic unit tests to complex optimizations. It’s not enough for the repository to be a database administrator; it must also fulfill the work of a housekeeper, a reviewer, a quality control expert, and even a policeman. Savvy development teams rely more on repositories to get their work done.

Hot 5: Robotic Process Automation

In the old days, you had to write some code to get anything done. Programmers needed to learn about variables, remember all the rules about types, scopes, and syntax, and then talk about rules about code quality, which usually came down to a statement about non-functional whitespace.

New tools like “robotic process automation” are changing the dynamic. However, this is only enhancing the data processing routine. Today, savvy non-programmers can do a lot of work using tools that eliminate much of the development process. Any user can work on a spreadsheet, there is no hassle with closures, and some nice interactive results can be produced.