Simple S3 Proxy for Squirrel Updates

This nodejs package implements a squirrel update service using Amazon S3 as a backend. Your builds and build metadata live in S3. As update requests arrive, the server checks a well-known S3 bucket and key for a latest.json file, and uses this data to decide whether to send 204 if the client is already up to date, or JSON based on latest.json if the client should update.

Installation

Clone this repo or download a release, set environment variables according to "Configuration" below, then yarn dev.

To make the server useful, you'll need to upload a latest.json to your channel bucket as specified by FILBERT_CHANNEL_BUCKET and FILBERT_CHANNEL_BUCKET. Here's a sample:

{
  "build": "fubar.dmg",
  "name": "Fubar Pro",
  "notes": "Warning: processed in a facility that also processes nuts.",
  "pub_date": "2016-09-14T12:34:56-00:00",
  "version": "0.1.2"
}

Filbert uses semver to compare the client's version string with the server's version string.

Sample requests and responses:

• curl -i 'http://localhost:3000/status/test/0.0.1': 200, because 0.0.1 has an update available.
• curl -i 'http://localhost:3000/status/test/0.1.2': 204, no update available.
• curl -i 'http://localhost:3000/status/test/0': 400, because 0 is not a semver.

When an update is available, the response includes a url built from the FILBERT_BUILD_BUCKET, FILBERT_BUILD_PREFIX, and the build from that channel's latest.json.

Configuration

cp .env.sample .env

Filbert uses environment variables for its configuration.

• FILBERT_PORT: The listener port, default 3000.
• FILBERT_BUILD_BUCKET: The S3 bucket where builds live.
• FILBERT_BUILD_PREFIX: The prefix under FILBERT_BUILD_BUCKET where builds live. This can be empty.
• FILBERT_CHANNEL_BUCKET: The S3 bucket where channel information lives.
• FILBERT_CHANNEL_PREFIX: The prefix under FILBERT_CHANNEL_BUCKET where each channel's latest.json lives. This can be empty.

For AWS credentials configure a $HOME/.aws/credentials, AWS environment credentials, or IAM profile with GetObject access to the appropriate S3 content.