The cloud has paved the way for automation and sophisticated apps. With these great advancements comes good and bad. We’re now able to update our code instantaneously. We’re able to deliver new features at a much faster speed than ever before. But this added complexity means that cloud optimization is harder than ever.
Research reveals that 80% of finance and IT leaders report that poor cloud financial management has hurt their business. 69% of users regularly overspend their cloud budget by 25 percent. For most enterprises, cloud optimization is not happening anywhere near as effectively as it needs to.
How did we get to this point? Before DevOps, everything – writing code, testing, tuning, and deployment – had to be done manually. Now, with DevOps, enterprises can write code with continuous build, continuous testing, and continuous deployment. This results in high velocity, but the tuning phase is no longer part of the assembly line. As more features are delivered, rather than conducting proper cloud optimization, people are overspending and overprovisioning applications to ensure the reliability of their services.
Tuning is no longer performed as part of the timeline. It is only done in a crisis. Companies try to utilize application management tools to help inform how they should tune. But a human is still needed to make changes for cloud optimization. However, there are vast amounts of configurations. A simple container application can have more than 225 trillion resources and basic parameter permutations.
The cloud has a sophisticated infrastructure that engineers really have to work to understand. Also, engineers have to be deeply familiar with the application and how it works, which requires expertise. Before Devops, a company could optimize or update code once a year. However, with Devops and constant updates to code, new configurations need to happen at lightning speed all the time.
The result? Performance tuning and cloud optimization are beyond the reach of human intelligence. The number of parameter settings and resource settings available in today’s modern app deployments means that true optimization is impossible without help from AI.
Microservices are now the preferred methodology for scaling applications. Instead of one microservice to tend to, there are now ten to hundreds of microservices in different languages that require different configurations with different optimal settings. Code changes have also increased by 100 times than before. There is a faster base in security, and other bug fixes that change the third party library and cannot be controlled. This can then sometimes cause significant performance impact.
Humans are not capable of continuous cloud optimization. It is too complex a task to manually tune trillions of configurations at lightning speed, round the clock. Companies are updating their codes and systems at speeds impossible to match manually. In order to achieve continuous optimization AI is essential.