The promise of observability in a complex distributed environment is much more important to the Site Reliability Engineering team or IT Operations. Without the use of proper tools, it can be tough to maintain such a system for the long term, especially when we are using an abstraction layer such as Istio on top of the application.
Traditionally, monitoring stacks, such as events, stack traces, log4j, ELK, Fluentd, Splunk, and so on for collecting metrics, incrementing counters, preparing histograms, and so on need to be embedded or instrumented in the application code. Some frameworks provide integration with the metrics system. This issue becomes complicated when polyglot applications are used and different sets of APIs need to be used to instrument and collect metrics.
In Istio, sidecar proxies have full control over the network traffic, and they...