Changelog

0.8.0 (2023-10-17)

Features

explain fork transition in readme (0ce9bbe)

Bug Fixes

correct default branch (f262778)
update gitignore file (15cabbb)
update ruby version and simplify dependency management (5b69e7e)

0.7.0

Add service specific AWS sdk gems
#79: Environment updates can complete with errors
#82: Added a line to force STDOUT output every line
#85: Improve package not found error message
#86: Fix to for inactive stack updates when instances are 0

0.6.6

Add support for specifying (and overriding) a stack policy for the CloudFormation resource stack. (contributed by @jlabrecque)
Fix issue where deployment hangs if resource stack update fails and stack is rolled back, deployment will now fail when the resource stack update fails. (contributed by @jlabrecque)
Add support for creating environment base on a elasticbeanstalk template. (contributed by @djpate)

0.6.5

#66: Making it possible to specify other accepted health states when deploying (@dziemid)

0.6.4

fixes #65 - aws driver should be able to detect CNAME prefixes from both regionalized and legacy EB domains

0.6.3

Wait until inactive environment status is ready before applying settings.

0.6.0

Use aws-sdk 2.0
Ruby 2.2 support
Robust error handling on CloudFormation resource stack provisioning

0.5.1

Fix issue: worker tier is picking up old version of aws-sqsd

0.5.0

Worker tier support

0.4.12

Update default solution stacks.
Rails generator generates SECRET_KEY_BASE for rails 4 app.

0.4.11

Make gem depend on aws-sdk-v1 until we put in support for both v1 and v2

0.4.10

Fix #33

0.4.9

Ignore AWS::ElasticBeanstalk::Errors::OperationInProgressFailure error when delete application version

0.4.8

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

0.4.7

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)

0.4.6

Make elasticbeanstalk event polling robust against clock shifting problem.

0.4.5

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)

0.4.4

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.

0.4.3

Backoff and retry when AWS::ElasticBeanstalk::Errors::Throttling happens during polling events

0.4.2

Allow provding different deploy strategy for different components.

0.4.1

Remove options for delete all environments ("-d --all"), because it is too destructive and not recoverable.
Experimental support for multiple components deployment.

0.4.0

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

0.3.9

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)

0.3.8

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.