August 2022 Summaries

AWS provides several load balancer types that can be used in conjunction with a Kubernetes ingress, including both transport-based layer 4 (L4) and application-based layer 7 (L7) options. The choice of load balancer type is crucial as it affects how traffic is routed to the Ingress Controller, and this decision depends on four key criteria: where to terminate TLS, how to manage certificates, how to handle cleartext redirection, and how to preserve client information. L7 load balancers are able to more intelligently decide how to route requests, but are typically less efficient, while L4 load balancers are typically very efficient but can be limited in the types of traffic they can route and how intelligently they are routing requests to the Ingress Controller. The choice of load balancer type also affects how TLS certificates are managed, how cleartext redirection is handled, and whether client information is preserved.

The article compares two popular API frameworks, REST and gRPC, highlighting their benefits, limitations, and potential use-cases. REST APIs are widely adopted in modern web development for building stateless, scalable, and reliable APIs. They rely on HTTP verbs to perform CRUD operations and follow design constraints such as uniform interface, decoupling the client and server, stateless interface, cacheability, layered architecture, and code on demand. On the other hand, gRPC is a universal, open-source, high-performance RPC framework that enforces scalability and performance for microservice architecture. It uses function calls to enforce client-server communication in microservices built with different programming languages and relies on protocol buffers (protobufs), streaming, and the HTTP/2 protocol for message transmission. gRPC APIs are modeled as a resource hierarchy and follow design constraints such as resource & service-oriented design, open-source, layered architecture, payload agnostic, flow control, metadata exchange, extensions as APIs, standard status codes. The article concludes by stating that both REST and gRPC API design models provide a way to connect loosely-coupled services in modern software development. While REST APIs are best suited for applications requiring high-speed HTTP message transfer iteration and cloud applications with flexible changes in workload, gRPC is ideal for building internal systems and real-time streaming in low-power, low-bandwidth connections due to its ability to handle multiplexing and two-way communications.