As some of you may know, I'm working on a collection of scripts for Switch to more easily interface with Amazon Web Services. I know cloud solutions have been slow to be adopted by our industry, so I put together a little simple example of how you can make a service like Sendfile, Hightail, or WeTransfer, for your organization for free (or damn close), using on Switch. Hoping that this will demonstrate S3’s simplicity as well as the value of SwitchAWS, and of course, Switch
.The Problem:
We want a system that will allow any user in our organization to send very large files to other people via email. This system should be able to handle files of around 5-10GB and the files should expire automatically after a certain amount of time so we do not rack up long term storage charges. The user must be able to send a single file, a set of files, or a folder.
Requirements:
Switch Client, Switch Scripting, AWS CLI
Solution:
The user drags the files they want to send into the Switch Client. They are then prompted with this Submit Point metadata screen:
Most of the fields above are optional. They fill out what they want and click submit.
Within a few moments, an email goes out to the recipient(s) specified, with me (as the sender) BCC’d.
And best of all, in 14 days, that file will be removed from the S3 bucket automatically.
Explanation:
You can view the documented flow here:
https://5177154f0ca9401dd7db4b20a02fa27 ... UliZWVncUU
Download the actual flow here:
https://drive.google.com/file/d/0B726GN ... sp=sharing
Once they click submit, the files are compressed. If a single compressed file is sent, it does not get compressed again.
The compressed set of files goes is sent through the S3 Upload configurator which uploads the file into the target S3 bucket; in my case: “hosted-file-send”. Once this is complete, the S3 object URL is appended to the job’s Private Data with the key “S3UploadLocation”. We will use this in our email.
The job passes through a “Mail send” configurator which sends a nicely templated HTML email with a direct link to the S3 object.
To handle automatic file expiration, we will leverage an S3 Lifecycle Rule. We simply click on our bucket (‘hosted-file-send’ in my case) in the AWS S3 console and in the Lifecycle tab, click “Add rule”. We then just set the files to permanently delete after X number of days. I chose 14.
Then, you just give it a name and enable it. I called mine “Delete after 2 weeks”.
Now all objects that are added to this bucket will be automatically deleted 14 days later. Easy, right? I mean, look how simple that flow is. That's all S3's simplicity that allows us to do this so easily.
If you're unfamiliar with S3 pricing, it's dirt cheap for an application like this. I think I calculated that a 5GB file to upload and get downloaded by the recipient and hosting it for a month would cost 42 cents. Well covered by their free tier...
Any questions, comments, improvements? Feel free to chime in.
