Example links
- https://itnext.io/on-grpc-load-balancing-683257c5b7b3
- https://kubernetes.io/blog/2018/11/07/grpc-load-balancing-on-kubernetes-without-tears/
- https://linkerd.io/2/getting-started/
Current issues:
- linkerd adds it self to all deployments via a webhook, we need to exclude linkerd from every namespace we dont want it, as by default it will add to all
- Modify the namespaces of the other applications to not apply the linkerd attribute to auto-apply
This repository contains a gRPC programming exercise which is based on the basic tutorial found in the official gRPC documentation.
The objective of this exercise is to explore the following gRPC features:
- Unary RPC
- Client streaming RPC
- Server streaming RPC
- Full duplex RPC
- Interceptors (to return faulty responses)
- Health checks
- Load balancing
- gRPC Metadata
It is developed using the following software:
- Minikube v0.33.1
- Go 1.10.2
- Protoc 3.6.1
- dep v0.5.0
- Linkerd2 edge-19.1.3
The gRPC server exposes the 4 interfaces:
| API | Description |
|---|---|
GetFeature |
Obtains the feature at a given position. |
ListFeatures |
Obtains the features available within the given rectangle, via server-side streaming. |
RecordRoute |
Accepts a stream of points from the client and returns a summary of the route traversed. |
RouteChat |
Accepts a stream of route notes from the client and returns another stream of notes to the client. |
It also uses interceptors to return faulty responses.
The gRPC client is used to make RPC calls to the server APIs. It can be started in two modes:
| Mode | Description |
|---|---|
FIREHOSE |
The client issues random calls to all 4 APIs in an infinite loop. |
REPEATN |
The client issues N calls to the selected API. The client will exit once it repeated N calls. |
To run the server and client locally:
$ make server
$ make client
Ensure that your local hostname resolves to 127.0.0.1. If not, add it to your /etc/hosts file.
To try out the client-side round robin load balancing, start multiple instances of the servers:
$ SERVER_PORT=8080 make server &
$ SERVER_PORT=8081 make server &
$ SERVER_PORT=8082 make server &
$ make client
Notice the logs of the servers as requests are received from the client.
To build the Dockerfile on Minikube:
$ make image
To meshed the server and client on Minikube with the Linkerd 2 proxy:
$ make mesh
To build the protobuf files:
$ make proto
To delete all resources from Minikube:
$ make clean