Imagine you're a DevOps engineer at a rapidly growing social media company, responsible for monitoring their flagship web application. The application, deployed on Managed Instance Groups behind an HTTP(S) load balancer, is generating logs at such a volume that the project is exceeding the logging API quota. Despite implementing exclusion filters in Cloud Logging, the issue continues. What could be the underlying problem?