goa VS gRPC

My experience of Golang is that dependency injection doesn't really have much benefit. It felt like a square peg in a round hole exercise when my team considered it. The team was almost exclusively Java/Typescript Devs so it was something that we thought we needed but I don't believe we actually missed once we decided to not pursue it.

If you are looking at OpenAPI in Golang I can recommend having a look at https://goa.design/. It's a DSL that generates OpenAPI specs and provides an implementation of the endpoints described. Can also generate gRPC from the same definitions.

We found this removed the need to write almost all of the API layer and a lot of the associated validation. We found the generated code including the server element to be production ready from the get go.

See https://goa.design/. It automates all the comms stuff, so you just write: 1) a design file showing your functions, 2) an implantation of those functions, and 3) a very generic "main.go" (basically the same for all your services) that decides "how is this exposed over gRPC or REST or other comms?". The rest of the code is generated.

Perhaps the easiest way to find out how to do something (especially when using Meta) is to search the test cases when you have cloned the source code.

If you really need a framework, you can take a look at Echo or, for a contract-first approach, https://goa.design/

Few folks in here are (rightly) frustrated with the code generation story and broader tooling support around the OpenAPI standard. I've found a few alternative approaches quite nice to work with:

- Use a DSL to describe your service and have it spit out the OpenAPI spec as well as server stubs. In other words, I wouldn't bother writing OpenAPI directly - it's an artifact that is generated at build time. As a Go user, I quite like Goa (https://goa.design/) but there are others shared in here like TypeSpec.

- There are situations where sticking a backend-for-frontend (BFF) in front of APIs can yield great productivity boosts. For example, in the past we built a thin GraphQL proxy that calls out to a poorly structured REST API. Integrating with that was much more convenient. Most recently, I've been playing with a BFF built with tRPC (https://trpc.io/) which calls out to a REST API. It seemed to provide an even better experience if you use TypeScript on the front-end and in the BFF. It does not have a codegen step and I was really pleased with how fast I could iterate with it - granted it was a toy project.

One of the biggest issues I see is that you are using the same models for API as you are for the database. That wouldn’t fly in a real work system. And even though your doing simple CRUD I would introduce another layer for business logic. You should never have the Controller calling you database code directly. It never “stays” that simplistic. One of the easiest ways to deal with this is to use Goa. https://goa.design/ It takes care of generating your API models and it creates the Interfaces to implement for your business logic. Furthermore it creates OpenAPI documentation (something missing in this design that is a must for commercial development).

I left PHP for Go.

- with http://sqlc.dev I don't have to write ORM or model code anymore.

- with http://goa.design I can have well-documented API's that any team can generate a client for in any language. It also generates the HTTP JSON and gRPC servers for me so I can focus on my logic.

- with https://github.com/99designs/gqlgen I can define GraphQL revolvers that play well with sqlc (any RDBMS) or I can use a key-value store.

- speaking of key-value stores, Go allows them to be embedded! Even SQLite now has the https://litestream.io/ project to make it super simple to use a durable, always backed-up SQLite database even in a serverless context.

Go is faster, uses less memory, and has really-well designed stdlib without all the bugs I used to face trying to use the PHP stdlib.

Goa and Kong are some of the best frameworks to develop and deploy microservices. They provide features such as out-of-the-box support for service discovery, routing and authentication that make it easier to build more complex applications. There are also newer architectural frameworks with less steep learning curves like GPTDeploy that lets you build and deploy microservices with a single command.

Or the one we use at work: https://goa.design/ Goa does a lot more and maybe more than you need. We use it as it can generate both REST and gRPC as well as API models and OpenAPI documentation (JSON and YAML).

gRPC, built on HTTP/2, inherently supports flow control. The server can push updates, but it must also respect flow control signals from the client, ensuring that it doesn't send data faster than what the client can handle.

Yes, grpc_cli tool uses essentially the same mechanism except implemented as a grpc service rather than as a stubby service. The basic principle of both is implementing the C++ proto library's DescriptorDatabase interface with cached recursive queries of (usually) the server's compiled in FileDescriptorProtos.

See also https://github.com/grpc/grpc/blob/master/doc/server-reflecti...

The primary difference between what grpc does and what stubby does is that grpc uses a stream to ensure that the reflection requests all go to the same server to avoid incompatible version skew and duplicate proto transmissions. With that said, in practice version skew is rarely a problem for grpc_cli style "issue a single RPC" usecases: even if requests do go to two or more different versions of a binary that might have incompatible proto graphs, it is very common for the request and response and RPC to all be in the same proto file so you only need to make one RPC in the first place unless you're using an extension mechanism like proto2 extensions or google.protobuf.Any.

While gRPC and Apache Thrift have served the microservice architecture well, CloudWeGo's advanced features and performance metrics set it apart as a promising open source solution for the future.

The loadBalancingConfig is what we use in order to decide which policy to go for (round_robin in this case). This JSON representation is based on a protobuf message, then why does the name resolver returns it in the JSON format? The main reason is that loadBalancingConfig is a oneof field inside the proto message and so it can not contain values unknown to the gRPC if used in the proto format. The JSON representation does not have this requirement so we can use a custom loadBalancingConfig .

The Dart implementation of gRPC which puts mobile and HTTP/2 first. It's built and maintained by the Dart team. gRPC is a high-performance RPC (remote procedure call) framework that is optimized for efficient data transfer.

gRPC is a high-performance, open-source RPC (Remote Procedure Call) framework initially developed by Google. It uses Protocol Buffers for serialization and supports bidirectional streaming.

In general, tunneling through HTTP2 turns out to be a great choice. There is a RPC protocol built on top of HTTP2: gRPC[1].

This is because HTTP2 is great at exploiting a TCP connection to transmit and receive multiple data structures concurrently - multiplexing.

There may not be a reason to use HTTP3 however, as QUIC already provides multiplexing.

I expect that in the future most communications will be over encrypted HTTP2 and QUIC simply because middleware creators can not resist to discriminate.

[1] <https://grpc.io>

Even in the https://grpc.io blog says this

gRPC