How do you scale network service mesh to support growing network demands and complexity?

1 Challenge 1: Service discovery

One of the main challenges for scaling NSM is how to discover and register network services across different clusters, clouds, or networks. Service discovery is the process of finding and identifying the available network services and their locations, attributes, and capabilities. Service discovery is essential for NSM to establish and maintain connections between network services, as well as to provide load balancing, fault tolerance, and security policies. However, service discovery can become complex and inefficient when the number and diversity of network services increase.

2 Solution 1: Federated service registry

One possible solution for scaling service discovery is to use a federated service registry, which is a distributed system that maintains a global view of the network services and their metadata. A federated service registry can synchronize and coordinate the local service registries of different clusters, clouds, or networks, and provide a unified interface for querying and accessing the network services. A federated service registry can also support dynamic updates, multi-tenancy, and encryption of the service metadata.

3 Challenge 2: Service communication

Another challenge for scaling NSM is how to enable secure and efficient communication between network services across different domains, regions, and protocols. Service communication is the process of exchanging data and messages between network services using various transport protocols, such as TCP, UDP, HTTP, or gRPC. Service communication is crucial for NSM to provide high-performance, reliable, and consistent connectivity between network services, as well as to support service-level agreements, observability, and tracing. However, service communication can become costly and complex when the latency, bandwidth, and security requirements vary across different network scenarios.

4 Solution 2: Service proxy

One possible solution for scaling service communication is to use a service proxy, which is a software component that acts as an intermediary between network services and handles the transport-level details of the communication. A service proxy can abstract away the heterogeneity and complexity of the underlying network infrastructure, and provide a uniform interface for the network services. A service proxy can also implement various features, such as encryption, authentication, load balancing, routing, caching, compression, and telemetry.

5 Challenge 3: Service orchestration

A third challenge for scaling NSM is how to coordinate and manage the lifecycle of network services across different clusters, clouds, or networks. Service orchestration is the process of automating and controlling the creation, deletion, update, and configuration of network services and their dependencies. Service orchestration is vital for NSM to provide agility, scalability, and reliability for the network services, as well as to support policy enforcement, resource allocation, and error handling. However, service orchestration can become difficult and error-prone when the number and complexity of network services increase.

6 Solution 3: Service controller

One possible solution for scaling service orchestration is to use a service controller, which is a software component that acts as a central authority for managing the network services and their state. A service controller can interact with the service registry, the service proxy, and the network infrastructure, and provide a declarative and consistent way of defining and applying the desired state of the network services. A service controller can also monitor and reconcile the actual state of the network services, and trigger actions or events based on the changes or events in the network.