From 5a9960f0b65ab90ec346f59f45b977d6375b4c77 Mon Sep 17 00:00:00 2001 From: Vishnu Bharathi Date: Tue, 1 Aug 2023 03:40:47 +0530 Subject: [PATCH] chore(readme): improve readme --- README.md | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 4821200..a7d16ef 100644 --- a/README.md +++ b/README.md @@ -17,13 +17,26 @@ waymond is - Built with Go. ## Motivation - -There are a lot of autoscalers out there like [AWS EC2 Autoscaling](https://docs.aws.amazon.com/autoscaling/ec2/userguide/what-is-amazon-ec2-auto-scaling.html), [kubernetes autoscalers](https://github.com/kubernetes/autoscaler), [nomad autoscaler](https://github.com/hashicorp/nomad-autoscaler), etc. Most of them are suited towards specific type of targets (example: nomad autoscaler could only be used for nomad). There is a good deal of overlap between those autoscalers in-terms of what and how an autoscale can happen. +Most of the auto-scalers nowadays are very specific to particular use-cases. They fall into one or more of the following categories: + +- Cloud specific + - [AWS EC2 Autoscaling groups](https://docs.aws.amazon.com/autoscaling/ec2/userguide/what-is-amazon-ec2-auto-scaling.html) + - [GCP Managed Instance groups](https://cloud.google.com/compute/docs/instance-groups) +- Platform-specific + - [Kubernetes autoscaler](https://github.com/kubernetes/autoscaler) + - [Nomad autoscaler](https://github.com/hashicorp/nomad-autoscaler) + - [Karpeneter](https://karpenter.sh/) + - [Keda](https://keda.sh/) +- Type and workload-specific + - Horizontal scaling + - Vertical scaling + - Compatible with containers only + - Compatible with VMs only + +There is a good deal of overlap between those autoscalers in-terms of what and how an autoscale can happen. Waymond tries to support a variety of autoscaling targets from operating system processes to kubernetes clusters and everything in-between like traditional VMs. One of the main goals of the project is to make it very easy to autoscale mixed type of targets. Truly anything and anywhere! The original idea for waymond came up while trying to autoscale CI/CD workloads in self-hosted infrastructure. Example: Autoscale the number of CI agents running as systemd processes inside a big EC2 VM and when we run out of limits there, try to bring up new EC2 VMs that run one CI agent per machine for a given CI job queue. At the sametime, autoscale the agents running in a kubernetes cluster when jobs are arriving in a different CI job queue. -waymond tries to support a variety of autoscaling targets from operating system processes to kubernetes clusters and everything in-between like traditional VMs. One of the main goals of the project is to make it very easy to autoscale mixed type of targets. Truly anything and anywhere! - ## Architecture ![architecture](https://user-images.githubusercontent.com/4211715/222922530-fda823c7-1a72-4156-99ac-3d249e4e8e47.png)