mirror of
https://github.com/verdaccio/verdaccio.git
synced 2024-11-13 03:35:52 +01:00
55 lines
1.7 KiB
Markdown
55 lines
1.7 KiB
Markdown
---
|
|
id: amazon
|
|
title: "Amazon Web Services"
|
|
---
|
|
|
|
This document describes several approaches for deploying Verdaccio in the AWS cloud.
|
|
|
|
## EC2 {#ec2}
|
|
|
|
[CloudFormation template for deploying this stack.](https://github.com/verdaccio/verdaccio/blob/master/contrib/aws/cloudformation-ec2-efs.yaml)
|
|
|
|
Architecture:
|
|
|
|
```
|
|
Clients
|
|
|
|
|
| (HTTPS)
|
|
v
|
|
Application Load Balancer
|
|
|
|
|
| (HTTP)
|
|
v
|
|
EC2 Auto Scaling Group (Amazon Linux 2)
|
|
Docker image (Verdaccio)
|
|
|
|
|
| (NFS)
|
|
v
|
|
Elastic File System
|
|
```
|
|
|
|
Architecture notes:
|
|
* Deploy this stack into the region closest to your users for maximum performance.
|
|
* We use an auto scaling group primarily for self-healing. The system requirements of Verdaccio are pretty low, so it's unlikely you'll need multiple instances to handle traffic load.
|
|
* Because Amazon Linux 2 doesn't include Node, we run Verdaccio as a Docker image rather than natively on the instance. This is faster and more secure than relying on third party package sources for Node.
|
|
* Elastic File System is cheap and stateful, and works across AZs. An alternative would be the [third-party S3 storage plugin](https://github.com/remitly/verdaccio-s3-storage).
|
|
* For backup, use AWS Backup
|
|
|
|
Estimated monthly cost for a small installation (in us-east-1):
|
|
* ALB (1 LCU average): $22.265/mo
|
|
* EC2 (t3.nano): $3.796/mo
|
|
* EBS (8gb): $0.80/mo
|
|
* EFS (5gb): $1.5/mo
|
|
* Data transfer: (10gb): $0.9/mo
|
|
* **TOTAL:** Under $30/mo
|
|
|
|
## ECS {#ecs}
|
|
|
|
You can deploy Verdaccio as a task with an [ECS Volume](https://docs.aws.amazon.com/AmazonECS/latest/developerguide/using_data_volumes.html) for persistent storage.
|
|
|
|
Note: Fargate doesn't support persistent volumes, so you have to use the S3 storage plugin.
|
|
|
|
## EKS {#eks}
|
|
|
|
See the documentation pages on [Kubernetes](kubernetes) and [Docker](docker).
|