Runtime logs are typically gathered by an agent interacting with the JVM, Node.js, or Python instrumentation APIs. These VMs provide a treasure trove of information regarding runtime systems. Metrics such as objects, memory allocation, threads in use, and DB connections can all be collected on a periodic basis and sent to a centralized system for monitoring.
In addition to the runtimes, these could also have the device heartbeats to know the status of the devices, and whether it is in working conditions or not can also be collected.
