- Wait until inactive environment status is ready before applying settings.
- Use aws-sdk 2.0
- Ruby 2.2 support
- Robust error handling on CloudFormation resource stack provisioning
- Fix issue: worker tier is picking up old version of aws-sqsd
- Worker tier support
- Update default solution stacks.
- Rails generator generates SECRET_KEY_BASE for rails 4 app.
- Make gem depend on aws-sdk-v1 until we put in support for both v1 and v2
- Fix #33
- Ignore AWS::ElasticBeanstalk::Errors::OperationInProgressFailure error when delete application version
- Raise an error if the environment launched with problems (contributed by kmanning)
- Add --stack-name option that let's use choose the name of the cloud formation stack to operate on (contributed by NET-A-PORTER)
- Document typo/grammar fix (contributed by stig)
- Retry on AWS API throttling error when operating on versions
- Added blue-only deployment strategy (a variant of blue-green) that skips the cname swap so that the newly deployed code remains on the inactive "blue" instance. (contributed by jlabrecque)
- Make elasticbeanstalk event polling robust against clock shifting problem.
- Rails 3+ support: Rails Generator to install default configurations with Postgres RDS resource and everything need for blue-green deployment
- Add ability to tag beanstalk environment (from @pmcfadden)
- Fix S3 upload on Windows.
- Experimental support for "inactve_settings" options, which can be used to automatically scale down the inactive environment.
- Update default solution stack name to latest supported JAVA stack.
- Validate solution_stack option before creating environment and list available solution_stacks if not valid.
- Backoff and retry when AWS::ElasticBeanstalk::Errors::Throttling happens during polling events
- Allow provding different deploy strategy for different components.
- Remove options for delete all environments ("-d --all"), because it is too destructive and not recoverable.
- Experimental support for multiple components deployment.
- revert back all changes from 0.3.8 to 0.3.9. Elasticbeanstalk haven't relex the real unique constain. The actually contrain is you can not have environment name cross different application
- Fix hang problem introduced in 0.3.8 when migrating old ElasticBeanstalk environment.
- Fix issue #13 (Deployment via S3 object broken since 0.3.7)
- Change ElasticBeanstalk environment name pattern. Stop using sufix hash to make eb environment global unique. (Because ElasticBeanstalk does not require environment has globally uniqe name any more.)
- Add migration logic so that if a ElasticBeanstalk environment with legcy name exists, eb_deployer will automaticly terminate it and replace it with ElasticBeanstalk environment has new name pattern.