How can you measure cloud architecture performance in real time?

Measuring a solution is important to provide a rational metrics that provides a common way for the entire team to estimate either success or failure of a project or initiative. It could be a quantitative or qualitative metric in most cases and they have their value to evaluate how close the designed solution meets its original objectives, and also why to evaluate modifications and its priorities in the future. I have seen many cases that measurements that may seem irrelevant at the beginning of a project to become gold mine down the road when there is enough historical data to learn from the past to have data-driven decision-making process and evaluation of lessons learned. It for certain can motivate the team when consciously explored.

Another aspect is always to target where to measure, closer to end-user/application experience or the back-end resources. Important to consider the metrics in the context of the traffic/data flow through the solution from start of the journey through the landing destination. Always important to start with a few but meaningful insights. Too few metrics may not capture the potential and important metrics, but on the other hand, too many metrics can cause distraction and diffusion of focus and major aspects to be kept in the main attention.

Measuring cloud architecture performance in real time depends on the platform you're using. For example, in AWS, you can utilize AWS CloudWatch, and in GCP, you can use Google Cloud Monitoring. These platforms offer specific tools and metrics to monitor factors like latency, throughput, resource utilization, error rates, and response time. You can also use a third-party application to do it and take action automatically.

Every tool will have pros and constraints. There is not such thing as one best tool. And any tool is better than no tool at all. Tools provided by the same vendor as the product, when available, provide quick adoption and easy integration with their own products to get you started. Open Source and third-party tool come with benefit of flexibility and holistic view and common metrics across different parts of a solution possibly coming from different providers in a singular experience way.

Make sure to align the metrics, by keeping at the top of mind what is the value to the business and project sponsors. It is also important to align the metrics with the original (or current) scope that the solution intends to address. Analyze the solution with metrics and tools is the art of telling the story of what it came to be, how it is performing and how it justifies its own cost or existence.

Be curious, find the right people with a different but constructive view and background. One of the great benefits of this new and connected online world is to meet and be exposed to the diversity, use it to your own constructive advantage.

A quick recent experience with a Hotel ERP implementation in Odoo customized for the client which is hosted on AWS. The ERP was been used at 5 different locations for managing the booking operations. On few days of the month, server went down because of CPU Burst Cycle exceeding the limit. Due to this we had to restart the instance to make the ERP up again. We monitored the pattern through AWS Cloud Watch as to when this issue occurs. We came to a conclusion that on specific days in a month when the flow of check-in and check-out increases and all the users of all locations performed the operation simultaneously the server couldn't handle the traffic. So we planned to upgrade.