https://github.com/fusakla/k8s-service-broadcasting

Tool allowing to broadcast/mirror/duplicate HTTP requests to all ready endpoints of a Kubernetes service.
https://github.com/fusakla/k8s-service-broadcasting

broadcast http kubernetes proxy traffic-mirroring

Last synced: about 1 year ago
JSON representation

Tool allowing to broadcast/mirror/duplicate HTTP requests to all ready endpoints of a Kubernetes service.

• Host: GitHub
• URL: https://github.com/fusakla/k8s-service-broadcasting
• Owner: FUSAKLA
• License: apache-2.0
• Created: 2020-01-26T20:24:32.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2020-01-27T22:59:04.000Z (over 6 years ago)
• Last Synced: 2025-04-14T22:56:06.925Z (about 1 year ago)
• Topics: broadcast, http, kubernetes, proxy, traffic-mirroring
• Language: Go
• Homepage:
• Size: 18.6 MB
• Stars: 6
• Watchers: 3
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md

Awesome Lists containing this project

README

          
# K8s service broadcasting

[![CircleCI](https://img.shields.io/circleci/build/github/FUSAKLA/k8s-service-broadcasting/master?style=flat-square)](https://app.circleci.com/github/FUSAKLA/k8s-service-broadcasting/pipelines?branch=master)

[![Docker Pulls](https://img.shields.io/docker/pulls/fusakla/k8s-service-broadcasting?style=flat-square)](https://hub.docker.com/repository/docker/fusakla/k8s-service-broadcasting)

Tool allowing to broadcast/mirror/duplicate HTTP requests to all ready endpoints of a Kubernetes service.

It supports two modes:

 - _Consistency_: Waits for response from all service endpoints and and response with one of erroneous response otherwise one of the succeeded. 

 - _Availability_: Responses with first successful response if any, otherwise with one of the error messages.

This behaviour can be controlled by the `--all-must-succeed` flag. Defaults to `true`.

## Usage

```bash

$ ./k8s-service-broadcasting --help

Tool allowing to broadcast/mirror/duplicate HTTP requests to all endpoints of Kubernetes service.

Waits for all of them to end and reports back failed request if any. If not returns last successful.

Usage:

  k8s-service-broadcasting [flags]

Flags:

      --all-must-succeed           By default if any backend fails, the whole request fails. If disabled one succeeded response is enough. (default true)

  -h, --help                       help for k8s-service-broadcasting

  -i, --interface string           Interface to listen on. (default "0.0.0.0:8080")

      --keepalive                  If keepalive should be enabled. (default true)

  -k, --kubeconfig string          Location of the kubeconfig, default if in cluster config or value of KUBECONFIG env variable. (default "/home/fusakla/.kube/conf/kubeconfig.yaml")

  -l, --log-level string           Log level (debug, info, warning, ...) default info. (default "info")

  -m, --metrics-interface string   Interface for exposing metrics. (default "0.0.0.0:8081")

  -n, --namespace string           Namespace to watch for.

  -p, --port-name string           Name of service port to sed the requests to.

  -s, --service string             Name of service to sed the requests to.

  -t, --timeout duration           Timeout for mirrored requests. (default 10s)

```

## Instrumentation

There is second interface for exposing Prometheus metrics and providing lifecycle endpoints.

This interface is running by default on `0.0.0.0:8081`. The endpointa are:

- `/metrics` Prometheus metrics

- `/-/healthy` liveness probe

- `/-/ready` readiness probe 

## Build

**single binary**

```bash

make build

```

## Deployment

See the [`kubernetes/`](./kubernetes) folder with example.