Upgrade to AWS SDK 3.0, reduce dependence to S3 specifically.
Support ERB in brillo.yaml
Support alternative primary keys to "id"
Fix bug where Brillo breaks when using the Octopus gem
Disable query cacheing to reduce memory growth over time.
Remove Caring specific scrubber code
Fix Postgres load failing because of existing connection
Fixed initializer. Brillo.configure
pre-initialization will now work.
Moved verification of config to post app initialization.
Support S3 config via your environment. To configure via the environment, just leave your Brillo.config.transfer_config.{access_key_id,secret_access_key}
blank.
Fix loading a postgres database with a password set.
Fix bug in S3 upload path. Files were previously uploading to their local system path in S3. I.e. bucket-name/my/rails/app/tmp/my-app-scrubbed.dmp.gz
.
Set us-east-1 as the default region, because this is S3s default and can be accessed from other regions.
Fixed postgres sequence not being set to MAX(id).
New
- BREAKING Brillo used to support loading your credentials from a YAML file at
/etc/ec2_secure_env.yml
but no longer does. It is now your responsibility to ensure the credentials are in the environment Brillo runs in. - Removed the dependency on the AWS Timkay CLI, instead using the AWS gem.
- Added support for configuring all S3 parameters in Ruby land
Fixed
- Fix mysql exec when no host specified
- Fix postgres exec when no host specified
- Fix postgres adapter reference
First public Brillo version!