CENPEC (Center for Studies and Research in Education, Culture and Community Action) is a non-profit organization that aims to develop projects, research, and methodologies that can improve the quality of public education in Brazil.
This organization acts as a partner in schools, public education environments, and in initiatives aimed at eradicating existing social inequalities. It develops actions that seek to contribute to the integral development of children and adolescents, in addition to the professional training of people working in education, thus expanding diversification and strengthening educational management.
CHALLENGE:
The client provides services across the country, mainly courses, which cannot become unavailable since many people access them daily. The main challenge was to maintain availability levels at 99%, as they were below 95%, making this a major challenge and the most important point to be resolved.
Also, we had to meet the needs at the lowest possible cost, since CENPEC is an NGO and has a limit on resources to be invested.
THE SOLUTION:
We used AWS EC2 Auto Scaling instances, along with load balancers and an RDS database. We left 1 instance running to save money. When necessary, it will scale up or down automatically, so that the client has their demands met with the savings they need.
RESULTS:
With the proposed solution, CodeBit managed to increase the availability of the CENPEC website by 4%. It also reduced costs by using only 1 instance with the right capacity, instead of a larger one with wasted resources.
With this, there were several benefits such as:
Greater fault tolerance: Amazon EC2 Auto Scaling can detect when an instance is not working properly, terminate it, and launch a new one to replace it.
Higher availability: Amazon EC2 Auto Scaling helps ensure that your application always has the right amount of capacity to handle the current traffic demand.
Better cost management: Amazon EC2 Auto Scaling can dynamically scale capacity up and down as needed. In other words, you pay for the instances you use. Thus, it guarantees greater savings, as instances are launched only when they are needed and terminated when they are not.




