s5cmd for very fast transfers

In case you have a fast connection of about 1-2Gbps and you want to utilize it for data transfers, you can use the s5cmd tool. It allows you to fully optimize the data transfer. The tool is available in form of compiled binaries for Windows, Linux and macOS. It is also available as a source code or docker image. Detailed information can be found on the project Github page.

Please insert into .aws/credentials the folowing options.

.aws/credentials
[default]
    aws_access_key_id = xxxxxxxxxxxxxxxxxxxxxx
    aws_secret_access_key = xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
    max_concurrent_requests = 200
    max_queue_size = 20000
    multipart_threshold = 128MB
    multipart_chunksize = 32MB

aws_access_key_id and aws_secret_access_key has been provided by admins while creating the S3 account.

Listing all buckets

s5cmd --endpoint-url=https://s3.clX.du.cesnet.cz ls

Simple file upload

s5cmd --endpoint-url=https://s3.clX.du.cesnet.cz cp myfile s3://bucket
ℹ️

How to achieve high transfer speed?
To achieve higher speed for data transfers it is necessary to modify the following parameters, particularly utilize or CPU cores and workers, see below.

s5cmd  --endpoint-url=https://s3.clX.du.cesnet.cz cp -c=8 -p=5000 /directory/big-file s3://bucket