wtf is service-mesh-performance?

Service Mesh Performance (SMP) is a specification that creates a common format for describing and comparing how different service meshes perform under various conditions. A service mesh is a dedicated infrastructure layer that handles communication between microservices in an application. When teams need to decide which service mesh to use, or how to tune their existing one, they need consistent, comparable data. That's what this project provides. The specification defines a structured way to capture several things together: the performance test setup, the service mesh configuration, the environment it runs in, the workload being tested, and the results. By putting all of these in a standardized format, someone can run a benchmark against one service mesh and meaningfully compare those results against another, apples to apples rather than apples to oranges. The project includes a sample results file so you can see what the output looks like in practice. The primary audience is platform engineers, DevOps teams, and architects who manage multiple microservices and need to evaluate or operate service meshes. For example, a team deciding between Istio and Linkerd could use SMP-formatted benchmarks to see how each handles latency or throughput under the same workload. The project ties into Meshery, an open-source management plane that works across many service meshes and serves as the reference implementation for the SMP spec. The project is being advanced through the CNCF's Service Mesh Working Group, which means it's part of a broader community effort to bring standardization to the service mesh ecosystem. It's worth noting that the README focuses on the specification itself rather than providing detailed setup instructions, it points to Meshery as the practical tool where the spec comes to life.