AWS CLI

The AWS CLI works with Hive Buckets. You need an access key.

Configure a profile

aws configure --profile dockhive
# AWS Access Key ID:     DH...
# AWS Secret Access Key: <your secret>
# Default region name:   auto
# Default output format: json

aws configure set s3.addressing_style path --profile dockhive

Point every command at the endpoint with --endpoint-url, or save it in the profile (AWS CLI 2.13 and later) so you can leave it out:

~/.aws/config

[profile dockhive]
region = auto
endpoint_url = https://s3.dockhive.io
s3 =
  addressing_style = path
  multipart_threshold = 64MB
  multipart_chunksize = 16MB

The examples below assume export AWS_PROFILE=dockhive and pass --endpoint-url for clarity.

Everyday commands

EP=https://s3.dockhive.io

# Upload a file and a folder
aws s3 cp ./report.pdf s3://my-app-uploads/reports/report.pdf --endpoint-url $EP
aws s3 cp ./images s3://my-app-uploads/images/ --recursive --endpoint-url $EP

# List
aws s3 ls --endpoint-url $EP                            # your buckets
aws s3 ls s3://my-app-uploads/reports/ --endpoint-url $EP

# Download
aws s3 cp s3://my-app-uploads/reports/report.pdf . --endpoint-url $EP

# Mirror a folder (only changed files are sent; --delete removes files that are gone locally)
aws s3 sync ./public s3://my-app-assets/ --delete --endpoint-url $EP

# Delete
aws s3 rm s3://my-app-uploads/reports/report.pdf --endpoint-url $EP
aws s3 rm s3://my-app-uploads/tmp/ --recursive --endpoint-url $EP

# Share a file for one hour
aws s3 presign s3://my-app-uploads/reports/report.pdf --expires-in 3600 --endpoint-url $EP

aws s3 cp and aws s3 sync switch to multipart uploads for large files automatically. Keep multipart_chunksize at 100 MB or less; the default of 8 MB is fine. See Limits.

Lower-level commands (s3api)

aws s3api exposes individual S3 operations:

# Object details and tags
aws s3api head-object --bucket my-app-uploads --key reports/report.pdf --endpoint-url $EP
aws s3api put-object-tagging --bucket my-app-uploads --key reports/report.pdf \
  --tagging 'TagSet=[{Key=team,Value=finance}]' --endpoint-url $EP

# Create a bucket (needs an Admin key for all buckets)
aws s3api create-bucket --bucket my-new-bucket --endpoint-url $EP

# Turn on versioning (needs an Admin key)
aws s3api put-bucket-versioning --bucket my-app-uploads \
  --versioning-configuration Status=Enabled --endpoint-url $EP
aws s3api list-object-versions --bucket my-app-uploads --prefix reports/ --endpoint-url $EP

More s3api examples are on the CORS, versioning, bucket policy and lifecycle pages.

Other S3 tools

Tools such as rclone, s3cmd, Cyberduck and restic work too. Use the same settings everywhere: endpoint https://s3.dockhive.io, region auto, path-style addressing, and your DH... access key.

Fast listings with rclone

By default rclone sends one extra request per file to read its content type and modification time, so listing thousands of files takes minutes. Add these flags and a listing of thousands of files takes a second or two:

rclone lsjson -R --fast-list --no-mimetype --use-server-modtime dockhive:my-bucket/backups/

--use-server-modtime uses the upload time Hive Buckets already returns in the listing. For rclone sync and rclone copy, use --fast-list and compare by size (--size-only) or checksum (--checksum) instead of modification time.