Releases: epam/aws-syndicate
Releases · epam/aws-syndicate
1.14.1
[1.14.1] - 2024-10-15
- Added support lambda layers with runtime DotNet
- Tagging added to
oas_v3 openapi
API Gateway. - SNS topic deletion fixed.
- Added an example of a lambda function with runtime
dotnet
- Fixed silent overwriting existing lambda with runtime Java during the command
syndicate generate lambda
1.14.0
[1.14.0] - 2024-08-28
- Changed deployment flow to work despite the latest deployment failed
- Changed deployment flow with the flag
--continue_deploy
to work despite the latest deployment being absent or succeeded - Implemented rolling back on error mechanism(flag
--rollback_on_error
) for deployment flow with the flag--continue_deploy
- Added support of lambda functions with runtime DotNet
- Added possibility to deploy/update resource-specific tags
- Added confirmation request mechanism for the
update
command in case the latest deployment failed - Added the flag
--force
for theupdate
command to run an update without confirmation request in case the latest deployment failed - Added proper messages for commands
update
andclean
if deployed resources are absent(output file not found) - Added logging of resource names that cause errors to improve error diagnostics
- Added enhanced logging of the
build
command execution - Added validation for existence of bundle and deploy names
- Added validation for incompatible parameters(
--events
,--resources
) of the commandsyndicate status
- Added the possibility to add resource-specific tags via the resource meta generation(flag
--tags
) - Added tags support to the Syndicate Java plugin. The
@Tags
and the@Tag
annotations added to the plugin. The Java plugin version updated to 1.14.0 - Changed the Java lambda template to use the Syndicate Java plugin version 1.14.0 for the new lambda generation.
- Event status added to the command
syndicate status
output - Reworked lambda triggers update to compare local event sources meta with the previous remote one
- Reworked lambda triggers deletion to not list every resource of the trigger type to remove it from lambda (EPMCEOOS-6112)
- The key
operation_status
inlatest_deploy
section of the syndicate state file(.syndicate) renamed tois_succeeded
- Fixed lock resolving issue
- Fixed an issue related to bucket name resolving in the s3_bucket policy
- Added logging of resource names that cause errors to improve error diagnostics
- Fix the resource update issue that occurs when a deploy_name is specified by user (not default one) but deployment
output for the latest deployment is empty - Fixed an issue where updating only certain resources caused the deployment output to be overwritten with only these
resources, instead of updating the existing meta - Fixed deployment failure if resource name is the same as resource type
- Fixed an issue related to resource name resolving if ARN is a list item
- Fixed an issue related to name resolving if one resource name contains another resource name
- Fixed an issue when a lambda deployment fails when a trigger defined in meta does not exist
- Fixed an issue related to updating lambda triggers
- Fixed lambda event source mapping inactivity after deployment
- Improved logging for the deletion of a resource that does not exist in the account
- Improved logging in case of the absence of resources to clean
- Fixed Lambda Layer packaging for the NodeJS runtime
- Fixed an issue where newly added resources (after deploy) were causing the update operation to fail
- Fixed a synchronization issue that prevented the batch job queue from being deleted before its state was fully updated
- Fixed an issue related to removing CloudWatch alarms that were used for Dynamodb autoscaling
- Fixed an issue with the absence of lambda's information in the output file in case deployment failed on the trigger configuration step
- Fixed an issue related to losing state after partial clean
- Fixed an issue related to the latest output resolving
- Fixed an issue related to the API Gateway throttling and cache configuration for ANY method
- Fixed an issue with
--force_upload
being required while syndicate assemble
1.13.1
[1.13.1] - 2024-08-05
- Speed up deletion of s3 bucket with lots of objects
Release 1.13
[1.13.0] - 2024-07-10
- Added possibility to configure
FunctionResponseTypes
for lambda functions - Updated maven plugin version to 1.12.0 with support of
FunctionResponseTypes
- Added possibility to set up Cognito user pool ID in lambda function environment variable
- Added possibility to set up Cognito user pool client ID in lambda function environment variable
- Fix lambda triggers deletion when removed from meta
- Fix resources dependencies resolving
- Fix losing successfully deployed resources from the output file during deployment with the option
--continue_deploy
- Fix API Gateway duplication in case of existing API Gateway with the same name
- Fix detection of usage
--rollback_on_error
option with an incompatible option--continue_deploy
- Changed datetime format for lock attributes in the
.syndicate
file to UTC format - The Syndicate Java plugin version updated to 1.13.0 with changes:
- The ResourceType enum for the @dependsOn annotation extended with new type ResourceType.COGNITO_USER_POOL
- The @EnvironmentVariable annotation for the Syndicate Java plugin improved to support the value transformer
- A new value transformer type created ValueTransformer.USER_POOL_NAME_TO_USER_POOL_ID
- A new value transformer type created ValueTransformer.USER_POOL_NAME_TO_CLIENT_ID
- The generate Java lambda template changed to use the Syndicate Java plugin version 1.13.0
Release 1.12.0
[1.12.0] - 2024-06-20
- Added ability for
clean
command to automatically resolve if--rollback
is needed. - Fixed an issue related to the
log group already exists
error while deploying or updatinglambda.
- The updated
syndicate deploy --continue_deploy
command can save the output and continue deploying resources. - Implemented the
rollback_on_error
flag for thesyndicate deploy
command. If the flag isTrue,
all resources deployed during the deployment process will be cleaned. - Fixed an issue related to deploying multiple resources of the same type. Now, it catches an
Exception
in case of a deployment error of one of the resources and returns it along with the outputs. - Fixed
deploy
command responses. - Added support for Python 3.12
- Fixed an issue related to ARNs resolving in case of empty resource name
- Fixed an issue related to improper filtering of resources in case of different types of filter usage
- Fixed an error related to SQS FIFO Queue availability regions
- Fixed an issue related to deploying SQS Queue with configured redrive_policy
- Fixed an issue when only the last s3 trigger was configured for the lambda
Release 1.11.6
[1.11.6] - 2024-05-24
- Added support of custom authorizer names in Open API specification security schemes
- Fixed quietness of errors while deploying/updating API Gateway via OpenAPI specification
- Fixed API GW deletion when openapi specification is invalid
- Fixed issue with the command
generate meta api_gateway_resource
- Fixed lambda function deployment fails in case of matching any resource name with prefix or/and suffix
Release 1.11.5
[1.11.5] - 2024-05-09
- Syndicate Java plugin patched to version 1.11.1 to exclude extra dependencies
- Fixed an error related to export OpenAPI specification in extended prefix mode
Release 1.11.4
[1.11.4] - 2024-04-30
- Added support of EC2 Launch Templates(
ec2_launch_template
) - Change log level (for non-last exceptions) in retry decorator from error to warning
- add
.syndicate-config-*/
to generated project gitignore - Fixed an error in case of an attempt to delete a nonexistent API Gateway
- Add error log in case of invalid resource type
- Fixed an error in case of an attempt to delete a nonexistent IAM Policy
- Fixed an error in case of an attempt to delete a nonexistent SNS Topic
- Fixed an error related to
UpdateEventSourceMapping operation
during SQS trigger deployment
Release 1.11.3
[1.11.3] - 2024-04-16
- Added support of Python 3.12 lambda runtime
- Root account removed from trusted entities of autogenerated lambda’s role
- Improve bundle name validation: exclude folder path from length validation
-
- Added an ability to build lambdas and lambda layers with local requirements (for
nodejs
runtime)
- Added an ability to build lambdas and lambda layers with local requirements (for
- Implemented subnet group deletion with deletion of the related DAX if the syndicate deployed it
- Fixed deploy target bucket key compound appending to artifact src in swagger ui resource
- Clarified the error message when copying NodeJS dependencies folder
- Changed the error log message to the warning when updating log group retention period
- Made a bunch of changes to the CloudWatch Alarm resource:
- added the ability to add ALARM actions for Lambdas and System Manager (incident manager);
- added new parameters to the resource description in the deployment_resources:
description
,datapoints
,dimensions
andevaluate_low_sample_count_percentile
; - added new parameters to the
syndicate generate meta cloudwatch_alarm
command:description
,datapoints
andevaluate_low_sample_count_percentile
; - added more values to the
comparison_operator
parameter:LessThanLowerOrGreaterThanUpperThreshold
,LessThanLowerThreshold
,GreaterThanUpperThreshold
- Fixed bug when lambda invocation permissions in lambda`s alias were not removed after the web_socket_api_gateway was destroyed
- Fixed displaying help messages with partially specified credentials
- Added more error logs in retry function which assumes aws credentials
- Added verbose mode (flag
--verbose|-v
) - improved project lock logic to make it configurable and time-limited
Release 1.11.2
[1.11.2] - 2024-03-28
- fix issue with syndicate not uploading deploy output on fail