I cannot give away too many details since this is a work thing, but I’m absolutely blown away by the volume of incoming log data that ElasticSearch is able to handle. We have a significant number of nodes sending 2-3 messages a second and hundreds of thousands of nodes. Want to “grep” all the logs? Write an ES query (which I’m about to do now). You can even setup deployments within the Kubernetes cluster to respond in any number of ways to stuff detected in the log queries and take proactive action to remediate. This is where stuff like Kubernetes starts to make sense. This type of thing would be a totally different beast in a VM only architecture.