razzo-lunare/s3

Fast simple s3 cli

S3

Dead simple and fast S3 sync

How is this faster?

Most S3 sync tools will call the list API for the root path you provide and rely on the S3 API to do the heavy lifting to list your objects. This can be really slow because it's bounded to one thread. Instead this tool recursively calls the S3 api using a large goroutine pool to process each subsequent directory until it list all your objects.

Getting started

1. Download cli from the Github release

2. Setup the s3 config

3. Test the cli following the examples below

Source or Destination list

s3://<BUCKET NAME>/<OPTIONAL PATH ON S3>: A location inside an S3 bucket.

example with a subpath inside the bucket

s3://dates-bucket/2021-02-10/

example with only the bucket

s3://dates-bucket/

filesystem://<AN ABSOLUTE OR RELATIVE PATH>: A location on your local filesystem

example relative path

filesystem://outputs/dates/

example absolute path

filesystem:///root/outputs/dates/

Examples

download from s3

s3 sync \
    --config configs/s3.yaml \
    --source s3://fortuna-stock-data-new/ \
    --destination filesystem://../fortuna-stock-data/

upload to s3

s3 sync \
    --config configs/s3.yaml \
    --source filesystem://../fortuna-stock-data/
    --destination s3://fortuna-stock-data-new/