heroku-cloudwatch-drain

README

heroku-cloudwatch-drain

A Heroku HTTPS log drain that stores logs in CloudWatch Logs.

Getting started

Heroku

Locally

Build and install:

$ go install

Run:

$ heroku-cloudwatch-drain

Configuration

See all available configuration flags:

$ heroku-cloudwatch-drain -h

The AWS configuration is picked up from the environment. For a full list of environment variables and other ways to configure the AWS region, credentials, etc., see the SDK Configuration page.

Exception reporting

Set the HONEYBADGER_API_KEY environment variable to report panics and errors to Honeybadger.

If you've deployed heroku-cloudwatch-drain to Heroku, see Honeybadger's Heroku specific instructions for more detailed information.

New Relic Integration

Set the NEW_RELIC_LICENSE_KEY environment variable to enable the New Relic integration. Optionally, you can set the NEW_RELIC_APP_NAME to customize the app name on New Relic.

Sending logs

Logs should be sent to this application, with the log group name as the URL path. For example, if the heroku-cloudwatch-drain is available at https://drain.example.com/, and you wish to collect logs under the log group name my-app, the log drain URL should be https://drain.example.com/my-app.

HTTP Basic Auth is supported and can be configured via CLI flags.

Both the CloudWatch Logs log group and log streams are created automatically as requests come in. A new and unique log stream is created for each process.

AWS IAM permissions

The IAM policy containing the minimum required permissions to run this is:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "logs:CreateLogGroup",
        "logs:CreateLogStream",
        "logs:PutLogEvents",
        "logs:PutRetentionPolicy"
      ],
      "Resource": [
        "*"
      ]
    }
  ]
}

Contributing

The govendor tool is used for managing the vendor directory.

Run tests:

$ govendor test +local