This plugin sends uploaded files to Amazon S3 instead of writing them to the server file system.
Resized images are properly supported.
Payload team supports an official cloud storage plugin, different from this one.
The main difference is that this plugin allows configuring collection logic on the collection itself.
Payload implementation requires to define collection-specific stuff from plugins inside the global payload configuration file, which is (imho) bad design.
npm install payload-s3-upload --legacy-peer-deps
Payload requires legacy-peer-deps
because of conflicts on React and GraphQL dependencies (see Payload docs).
import { S3Client } from '@aws-sdk/client-s3';
import { buildConfig } from 'payload/config';
import s3Upload from 'payload-s3-upload';
export default buildConfig({
// ...
plugins: [
s3Upload(new S3Client({
region: process.env.AWS_REGION,
credentials: {
accessKeyId: process.env.AWS_KEY,
secretAccessKey: process.env.AWS_SECRET,
},
})),
],
});
import { S3UploadCollectionConfig } from 'payload-s3-upload';
const Media: S3UploadCollectionConfig = {
slug: 'media',
upload: {
staticURL: '/assets',
staticDir: 'assets',
disableLocalStorage: true,
s3: {
bucket: 'my-bucket',
prefix: 'images/xyz', // files will be stored in bucket folder images/xyz
// prefix: ({ doc }) => `assets/${doc.type}`, // dynamic prefixes are possible too
commandInput: {
// optionally, use here any valid PutObjectCommandInput property
// https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/clients/client-s3/interfaces/putobjectcommandinput.html
ACL: 'public-read',
},
},
adminThumbnail: ({ doc }) =>
`https://my-bucket.s3.eu-west-3.amazonaws.com/images/xyz/${doc.filename}`,
},
// create a field to access uploaded files in s3 from payload api
fields: [
{
name: 'url',
type: 'text',
access: {
create: () => false,
},
admin: {
disabled: true,
},
hooks: {
afterRead: [
({ data: doc }) =>
`https://my-bucket.s3.eu-west-3.amazonaws.com/images/${doc.type}/${doc.filename}`,
],
},
},
],
};
export default Media;
This plugin automatically uploads image variants in S3.
However, in order to retrieve correct URLs for the different sizes in the API, additional hooks should be implemented.
import { S3UploadCollectionConfig } from 'payload-s3-upload';
const Media: S3UploadCollectionConfig = {
slug: 'media',
upload: {
// ...
imageSizes: [
{
name: 'thumbnail',
width: 400,
height: 300,
crop: 'center'
},
{
name: 'card',
width: 768,
height: 1024,
crop: 'center'
},
{
name: 'tablet',
width: 1024,
height: null,
crop: 'center'
}
],
adminThumbnail: 'thumbnail',
},
hooks: {
afterRead: [
({ doc }) => {
// add a url property on the main image
doc.url = `${myBucketUrl}/${doc.filename}`
// add a url property on each imageSize
Object.keys(doc.sizes)
.forEach(k => doc.sizes[k].url = `${myBucketUrl}/${doc.sizes[k].filename}`)
}
]
},
fields: []
};
export default Media;
Please refer to the test files!