hubot-s3-brain

Store the hubot brain in S3.

This is a lightly modified version of hubot-scripts/s3-brain.coffee by IrishStyle. It has been updated to use the AWS S3 SDK package.

Take care if using this brain storage with other brain storages. Others may set the auto-save interval to an undesireable value. Since S3 requests have an associated monetary value, this script uses a 30 minute auto-save timer by default to reduce cost.

Installation

In hubot project repo, run:

npm install hubot-s3-brain --save

Then add hubot-rules to your external-scripts.json:

[
  "hubot-s3-brain"
]

Configuration

S3-brain can be configured with environment variables:

• HUBOT_S3_BRAIN_ACCESS_KEY_ID Optional AWS Access Key ID with S3 permissions
• HUBOT_S3_BRAIN_SECRET_ACCESS_KEY Optional AWS Secret Access Key for ID
• HUBOT_S3_BRAIN_BUCKET Bucket to store brain in
• HUBOT_S3_BRAIN_FILE_PATH Optional path/file in bucket to store brain at
• HUBOT_S3_BRAIN_SAVE_INTERVAL Optional auto-save interval in seconds
• HUBOT_S3_BRAIN_ENDPOINT Optional alternative S3 API endpoint

It's highly recommended to use an IAM account explicitly for this s3-brain. A sample S3 policy for a bucket named Hubot-Bucket would be:

{
   "Statement": [
     {
       "Action": [
         "s3:DeleteObject",
         "s3:DeleteObjectVersion",
         "s3:GetObject",
         "s3:GetObjectAcl",
         "s3:GetObjectVersion",
         "s3:GetObjectVersionAcl",
         "s3:PutObject",
         "s3:PutObjectAcl",
         "s3:PutObjectVersionAcl"
       ],
       "Effect": "Allow",
       "Resource": [
         "arn:aws:s3:::Hubot-Bucket/brain-dump.json"
       ]
     }
   ]
 }