I’m looking for a way to aggregate many relatively sparse signals into an aggregated stream, which has a higher update frequency than the individual streams. For example, let’s imagine we are measuring the RAM usage of a cluster of machines every 5 minutes, and I want to graph the total memory used by the entire cluster per second - in my case it’s OK to take the last measured usage for each machine as the “current” usage, but it would be great if I could set a TTL for its value, in case the server goes down and stops sending data. Python or SQL would be preferred 🙂 Thanks!