Back in the day, monitoring was simple. Proprietary monitoring agents sent infrastructure level metrics like CPU-usage and another agent sent logs to a syslog server to trawl for error messages.
However, as applications moved from virtual machine-based monoliths to cloud native microservice architectures running in the cloud, understanding the application and system state became much more difficult, requiring a more flexible approach to monitoring. Prometheus, the CNCF-backed open source metrics platform, has grown into the de-facto standard for monitoring cloud native environments, and we’ve moved from proprietary agents to open standards with many software components being pre-instrumented for quick and easy integration. Its popularity is, among other things, due to its inherent flexibility to explore data at will, arbitrarily querying the data to show different slices and views of data as teams investigate issues.
