Observability in Kubernetes deployment


Hi guys,

I’m currently trying to get a lagom based system ready for production and looking at ways to hook the microservices up to tracing, metrics, and logging system.

We nailed down most of the details for logging. As for tracing and metrics, initially Istio looked like a decent option that provides them along with service mesh, until we realized that there’s some requirements on service and port naming. Also the fact that lagom/lightbend are pretty opinionated when it comes to service discovery and deployment (there’s even a built in circuit breaker).

My question is, are service mesh like Istio/Linkerd or even Envoy recommended for lagom deployment? Or it’s better to hook everything up separately (hystrix/jaeger/prometheus/etc).