How would we be able to get solid evidence that we have an overloaded APM cluster / cloud proxy?
Cloud Proxy was running into memory issues. APM Cluster was fine. Added more pods and a new load balancer. Increased max memory on some cloud proxies.