servicemmesh

Servicemmesh is a term used to describe a dedicated infrastructure layer for managing communication between microservices in a distributed application. Note: the term is not standard; the canonical concept is commonly referred to as a service mesh. A servicemmesh provides features such as service discovery, load balancing, encryption, observability, and traffic management across service-to-service calls.

The architecture typically separates the system into a data plane and a control plane. The data plane consists of lightweight sidecar proxies injected beside each service instance, while the control plane configures the proxies, enforces policies, and manages identities and certificates. Proxies handle request routing, retries, timeouts, and fault injection, while the control plane distributes configuration, performs policy decisions, and handles certificate management for mutual TLS.

Key features include traffic management with fine-grained routing, canary releases, and blue-green deployments; security with mutual TLS, service identity, and access control; observability through metrics, distributed tracing, and centralized logs; and resilience features such as retries, circuit breakers, and timeouts.

Common implementations include Istio, Linkerd, Consul Connect, Kuma, and Open Service Mesh. These projects typically run on container orchestrators such as Kubernetes, but many support hybrid environments and VM workloads as well.

Limitations include added operational complexity, potential performance overhead, and the need for specialized skills to design and maintain policies and monitoring. See also: service mesh, sidecar pattern, mTLS.