Add Support for Amazon AWS S3 Service as a Storage Target

[Wayne Anderson]

Add support for Amazon Web Service (AWS) S3 storage buckets as a file transfer target for FileZilla.

My search would indicate that this request was last addressed in tickets 2741 and 2648, which at the time the recommended path was requesting Amazon support FTP.

In the intervening timeframe, the userbase of AWS S3 has exploded, there have existed a number of architectural reasons why API based access provides significant advantage. The present climate of FileZilla's use would establish a case for the interface metaphor so well liked in FileZilla to target S3 storage buckets as well.

The developer documentation, including the S3 API and developer guides is available at: ​http://aws.amazon.com/documentation/s3/

[Wayne Anderson]

Prior user comment not appropriate in nature, makes gratuitous changes.

Restoring ticket settings.

[Wayne Anderson]

Prior user comment not appropriate in nature, makes gratuitous changes.
Restoring ticket settings.

[Wayne Anderson]

Prior user comment not appropriate in nature, makes gratuitous changes.
Restoring ticket settings.

[Tim Kosse]

Is there a reference implementation of the S3 protocol so that I can run my own S3 server on my own hardware for testing?

[Wayne Anderson]

There is not a "reference implementation" per se, in that the toolkit is proprietary to Amazon's platform of inter-related AWS cloud services. Thus testing would be against a live instance on Amazon.

The positive news here is that Amazon has created an "AWS Fee Usage Tier" for just such testing scenarios where you can signup for a free AWS account and then are granted a minimum set of resources at no cost. You can then use the API to create and modify storage buckets, and store up to 5GB of data with up to 20k get requests and 2k put requests without charge.

This should be sufficient for basic development and testing scenarios. My experience developing for such things would indicate that if you are going to "bust" anything in the free tier, its going to be the transaction volumes testing multi-part upload for gross file speeds. Additional requests are charged at a single penny per 1,000 upload requests or 10,000 download requests, which should be very economical for continued testing against the live APIs.

The Free Usage Tier: ​http://aws.amazon.com/free/
The AWS S3 documentation: ​http://aws.amazon.com/documentation/s3/

[Adam Nielsen]

There are now a few open source server implementations that you can run locally instead of using Amazon's free services:

​https://github.com/scality/S3
​https://github.com/minio/minio

These two are compatible enough that you can use Amazon's official client tools to work with your local S3 instance.

Since the S3 protocol is built on standard HTTP requests, hopefully it won't be too much effort to add to Filezilla!

[Tim Kosse]

FileZilla Pro has support for S3.