Skip to main content

Fast uploader to S3

Project description

s3peat is a Python module to help upload directories to S3 using parallel threads.

The source is hosted at


s3peat can be installed from PyPI to get the latest release. If you’d like development code, you can check out the git repo.

# Install from PyPI
$ pip install s3peat

# Install from GitHub
$ git clone
$ cd s3peat
$ python install

Command line usage

When installed via pip or python install, a command called s3peat will be added. This command can be used to upload files easily.

$ s3peat --help
usage: s3peat [--prefix] --bucket [--key] [--secret] [--concurrency]
      [--exclude] [--include] [--dry-run] [--verbose] [--version]
      [--help] directory

positional arguments:
  directory            directory to be uploaded

optional arguments:
  --prefix , -p        s3 key prefix
  --bucket , -b        s3 bucket name
  --key , -k           AWS key id
  --secret , -s        AWS secret
  --concurrency , -c   number of threads to use
  --exclude , -e       exclusion regex
  --include , -i       inclusion regex
  --private, -r        do not set ACL public
  --dry-run, -d        print files matched and exit, do not upload
  --verbose, -v        increase verbosity (-vvv means more verbose)
  --version            show program's version number and exit
  --help               display this help and exit


$ s3peat -b my/bucket -p my/s3/key/prefix -k KEY -s SECRET my-dir/


This library is based around boto. Your AWS Access Key Id and AWS Secret Access Key do not have to be passed on the command line - they may be configured using any method that boto supports, including environment variables and the ~/.boto config..

Example using environment variables:

$ export AWS_ACCESS_KEY_ID=ABCDEFabcdef01234567
$ export AWS_SECRET_ACCESS_KEY=ABCDEFabcdef0123456789ABCDEFabcdef012345
$ s3peat -b my/bucket -p s3/prefix -c 25 some_dir/

Example ``~/.boto`` config:

# File: ~/.boto
aws_access_key_id = ABCDEFabcdef01234567
aws_secret_access_key = ABCDEFabcdef0123456789ABCDEFabcdef012345

Including and excluding files

Using the --include and --exclude (-i or -e) parameters, you can specify regex patterns to include or exclude from the list of files to be uploaded.

These regexes are Python regexes, as applied by, so if you want to match the beginning or end of a filename (including the directory), make sure to use the ^ or $ metacharacters.

These parameters can be specified multiple times, for example:

# Upload all .txt and .py files, excluding the test directory
$ s3peat -b my-bucket -i '.txt$' -i '.py$' -e '^test/' .

Doing a Dry-run

If you’re unsure what exactly is in the directory to be uploaded, you can do a dry run with the --dry-run or -d option.

By default, dry runs only output the number of files found and an error message if it cannot connect to the specified S3 bucket. As you increase verbosity, more information will be output. See below for examples.

$ s3peat -b my-bucket . -e '\.git' --dry-run
21 files found.

$ s3peat -b foo . -e '\.git' --dry-run
21 files found.
Error connecting to S3 bucket 'foo'.

$ s3peat -b my-bucket . -e '\.git' --dry-run -v
21 files found.
Connected to S3 bucket 'my-bucket' OK.

$ s3peat -b foo . -e '\.git' --dry-run -v
21 files found.
Error connecting to S3 bucket 'foo'.
    S3ResponseError: 403 Forbidden

$ s3peat -b my-bucket . -i 'rst$|py$|LICENSE' --dry-run
5 files found.

$ s3peat -b my-bucket . -i 'rst$|py$|LICENSE' --dry-run -vv
Finding files in /home/s3peat/ ...


5 files found.

Connected to S3 bucket 'my-bucket' OK.


s3peat is designed to upload to S3 with high concurrency. The only limits are the speed of your uplink and the GIL. Python is limited in the number of threads that will run concurrently on a single core.

Typically, it seems that more than 50 threads do not add anything to the upload speed, but your experiences may differ based on your network and CPU speeds.

If you want to try to tune your concurrency for your platfrom, I suggest using the time command.


$ time s3peat -b my-bucket -p my/key/ --concurrency 50 my-dir/
271/271 files uploaded

real 0m2.909s
user 0m0.488s
sys  0m0.114s

Python API

The Python API has inline documentation, which should be good. If there’s questions, you can open a github issue. Here’s an example anyway.


from s3peat import S3Bucket, sync_to_s3

# Create a S3Bucket instance, which is used to create connections to S3
bucket = S3Bucket('my-bucket', AWS_KEY, AWS_SECRET)

# Call the sync_to_s3 method
failures = sync_to_s3(directory='my/directory', prefix='my/key',
    bucket=bucket, concurrency=50)

# A list of filenames will be returned if there were failures in uploading
if not failures:
    print "No failures"
    print "Failed:", failures



Released November 20th, 2014.


Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

s3peat-0.4.7.tar.gz (9.7 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page