By Ross Schibler Co-Founder & CEO, Opsani

Most modern enterprises are aware that optimizing cloud applications is a must. Without optimization, money will be wasted and performance will be suboptimal. However, while the majority of business organizations have optimization strategies in place, the question is: Are current methods the most effective and efficient?

In short: Mostly no. Manual tuning remains the backbone of many companies' optimization strategies. However, the emergence of DevOps practices enabled rapid delivery of software, and we now ship code at a furious pace. The reality is that, because of this pace, manual tuning has become woefully ineffective. Manual optimization is inefficient, mundane, and wasteful. It produces suboptimal performance and lower availability.

In the modern era, the only way to guarantee performance and efficiency is through AI-powered, continuous optimization. This technology can empower employees to stop wasting time on mundane tasks; optimization can be handed off to machines, and staff can focus on more productive tasks.

Hidden Inefficiencies

Public cloud use has skyrocketed since the pandemic began. According to analyst firm IDC, cloud revenues have totaled $233.4 billion, a 26 percent year over year increase. Ergo, the effective optimization of cloud applications is now more important than ever. Are enterprises' practices working?

To measure and validate the impact of cloud optimization on SaaS companies, we conducted a month-long poll in which 1000 C-level executives gave their thoughts and input on how the spike in cloud services use has affected their organizations' ability to deliver the best user experience of their products and services, for the lowest costs. Companies surveyed ranged in size from 250 to 5000 employees.

When asked how often their organizations optimized their application stack, 82 percent of poll respondents said "regularly." 91 percent of respondents were "highly confident" or "confident" that their cloud applications were running efficiently.

Sounds pretty good, right? However, a thorough look at the data suggests a different reality. The survey points to processes and challenges that make the task of cloud optimization near impossible without cutting-edge technology.

Optimizing Cloud Apps: Beyond Human Power

The pandemic-triggered surge in cloud usage forced modern enterprises to rush the adoption of containers. In this environment, avoiding cost creep requires optimizing applications environments continuously. Manually, this continuous approach is impossible.

Consider the complexity of even a fairly simple application like Google Online Boutique, a 10-tier microservices application with a backend database. Even this application - a web-based e-commerce app where users can browse items, add them to a cart, and purchase them - presents serious challenges.

To determine the optimal performance of Google Online Boutique, 75 quintillion different configuration permutations have to be analyzed. As a point of reference, there are about 7.5 quintillion grains of sand on planet earth.

Adding to this complexity is the frequency of software updates. In our survey, respondents were asked how often their organizations released new code. 14.55 percent said hourly sprints; 43 percent indicated daily updates; and 37 percent released new software on a weekly basis.

Given the complexity of cloud apps, even those 91 percent of respondents who said they were confident of their apps' performance likely still have significant performance improvements and costs savings. When software is being pushed into production this quickly, the complexity is simply too much for traditional optimization techniques to handle.

AI and ML: Critical Optimization Elements

If the complexities of the optimization processes are beyond human capability, what's the answer? Artificial intelligence and machine learning. These are a requirement to achieve continuous and effective cloud app optimization. Many leaders know this; in our poll, 35 percent of respondents indicated they felt AI would enhance the output of their employee assets.

The good news is that business executives and leadership personnel are becoming more aware of how crucial artificial intelligence and machine learning have become. Our survey revealed that  80.6 percent of respondents leverage AI tools to speed up their decision-making purposes; 15 percent reported they did not have those tools yet, but were planning to implement them.

Once niche technologies, in recent years AI and ML have become increasingly widespread. In the context of optimization, the deployment of said technologies ensures faster discovery and implementation of the ideal configurations of cloud infrastructure, apps, processes, resources, and more.

AI-driven optimization eliminates the need to assign employees to perform manual tuning. This addresses one of the key concerns that came out in our poll. 54 percent of our respondents indicated that ‘employees wasting time on mundane tasks' was top of mind. 43 percent said their legacy processes were not reflective of today's technology landscape; 47.6 percent were having difficulty finding the talent needed to complete mission critical tasks.

AI-driven optimization addresses a range of other issues that matter to executives. In our poll:

• 65 percent of respondents said 'staying within a certain budget' was a top priority.
• 62 percent indicated maximizing resources, such as CPU, bandwidth, etc., was important to them.
• 62 percent cited the importance of 'making sure cloud apps maintain uptime'.
• 63 percent were focussed on the metric of transactions per second in their SLAs.
• 47 percent were focussed on latency, and 43 percent were concerned to optimize dollar transaction per second.
  

By engaging with quintillions of permutations on a real-time basis, AI-driven optimization addresses all of these needs.

An Optimized Future

Our poll reveals that people believe their optimization efforts are basically effective. However, they also worry about wasting human resources, and are concerned about ensuring they meet SLAs. These concerns are understandable, but they can be alleviated if leaders accept that current optimization efforts are dated, and embrace new technologies.

SaaS companies must understand that to ensure optimal performance of their apps, they can't rely on manual tuning any longer. Optimization has to be immediate and accurate. It has to achieve SLA standards yet still keeps cost manageable. It needs to leverage both AI and ML.

When AI and ML take the difficult and mundane job of optimizing the infrastructure, staff are unshackled, allowing them to perform tasks that add value to organizations. This is the future of how enterprises approach cloud applications.

##

To learn more about containerized infrastructure and cloud native technologies, consider joining us at KubeCon + CloudNativeCon NA Virtual, November 17-20.

About the Author

Ross Schibler Co-Founder & CEO, Opsani

Ross's idea to use AI and machine learning to automate continuous optimization for complex infrastructures came in a flash of insight-a contemporary solution for a contemporary challenge.

Before founding Opsani, Ross founded Topspin Communications and saw its success through to its acquisition by Cisco. Prior to this, Ross founded Rapid City Communications later acquired by Nortel, where the technology and product line was responsible for $2B in revenue.