fix: [CDS-115382]: ff cleanup CDS_ECS_STEADY_STATE_CHECK_ON_RETRIES

Author: satendra-harness

docs/continuous-delivery/cd-integrations.md

@@ -31,7 +31,6 @@ To enable a feature flag in your Harness account, contact [Harness Support](mail
 | PIPE_GITX_DISABLE_SKIP_GIT_WEBHOOK_REGISTRATION | Enables users to register GitX webhooks without creating them in Git providers, allowing webhook setup via an intermediary service without requiring connector-level registration permissions. **This feature is in Limited GA.** |
 | CDS_DEPLOY_SERVICE_IN_PARALLEL_FIX | Enables users to retain the selected chart version when toggling the *Deploy Service in Parallel* option, even when both are configured as runtime inputs. **This feature is in Limited GA**. |
 | PIPE_DISABLE_PIPELINE_NOTIFICATIONS_ON_ROLLBACK | Enables users to suppress independent `PIPELINE_START` and `PIPELINE_SUCCESS` notifications during rollback, ensuring that only the main pipeline's `PIPELINE_FAILED` notification is sent. **This feature is in Limited GA**. |
-| CDS_ECS_STEADY_STATE_CHECK_ON_RETRIES | Enables users to perform a health check on retry in the Fargate deployment EcsUpgradeContainer step, ensuring the step only succeeds if the service is healthy. **This feature is in Limited GA.** |
 | CDS_K8S_NOT_ADD_TRAFFIC_ROUTING_TO_CANARY_WORKLOAD | Enables users to prevent Virtual Services from being deleted during the Canary Delete step in Kubernetes deployments, preserving traffic routing configurations during rollback. **This feature is in Limited GA**. |
 | CDS_ASG_MULTI_DEPLOY_ROLLBACK_SUPPORT | Enables users to [deploy multiple ASGs in a single stage](/docs/continuous-delivery/deploy-srv-diff-platforms/aws/asg/asg-tutorial/#multi-asg-deployments-and-parallel-rollback-single-serviceenvironmentinfrastructure) for the same service, environment, and infrastructure, with automatic parallel rollback triggered on failure. **This feature is in Limited GA.** |
 | PIPE_FAIL_ALL_FAILURE_STRATEGY | Enables users to configure [fail fast behavior](/docs/platform/pipelines/failure-handling/fast-fail/) in parallel executions, where a failure in any step or stage immediately fails the pipeline and triggers the configured failure strategy. **This feature is in Limited GA**. |

release-notes/continuous-delivery.md

@@ -465,7 +465,6 @@ Harness introduced a series of user experience improvements to the GitOps Agent
 - Previously, the Canary deployment stage failed with a NullPointerException when attempting to access getPreviousReleaseVersion() on a null output from the Helm Canary step. This issue is resolved. (**CDS-111738, ZD-87118**) 
 - Previously, pipeline rollbacks triggered separate PIPELINE_START and PIPELINE_SUCCESS notifications for the rollback execution, followed by a PIPELINE_FAILED notification for the main pipeline. This behavior exposed rollback pipelines as independent executions. The issue is resolved and behind a feature flag `PIPE_DISABLE_PIPELINE_NOTIFICATIONS_ON_ROLLBACK`. Please contact [Harness Support](mailto:[email protected]) to enable this feature flag. (**PIPE-28097**)
 - Previously, the CD Git Clone step failed even when valid repository details were provided. This was due to an internal mapping error where the step failed with the message: *Step `[<step_name>]`* should map to single port. The issue is resolved. (**PIPE-28355, ZD-87095**)
-- Previously, the Fargate deployment EcsUpgradeContainer step did not perform a health check on retry, which led to the step being marked as successful even when the service was not healthy. This caused the pipeline to proceed under the false assumption that the deployment succeeded. The issue is resolved with a feature flag `CDS_ECS_STEADY_STATE_CHECK_ON_RETRIES`. Please contact [Harness Support](mailto:[email protected]) to enable this feature flag. (**CDS-111777, ZD-86907**)
 - Previously, the Canary Delete step in Kubernetes deployments incorrectly deleted Virtual Services as part of the workload during rollback. This led to loss of traffic routing configurations and potential service downtime. The issue is resolved with a feature flag `CDS_K8S_NOT_ADD_TRAFFIC_ROUTING_TO_CANARY_WORKLOAD`. Please contact [Harness Support](mailto:[email protected]) to enable this feature flag. (**CDS-111916, ZD-87449**)
 
 ### Version 1.95.5

release-notes/self-managed-enterprise-edition.md

@@ -1331,7 +1331,6 @@ gsutil -m cp \
 - Fixed an issue where the Canary deployment stage failed with a NullPointerException due to a null output from the Helm Canary step. [CDS-111738, ZD-87118]
 - Fixed an issue where rollback pipelines triggered separate PIPELINE_START and PIPELINE_SUCCESS notifications followed by a PIPELINE_FAILED notification for the main pipeline. This is resolved and behind the feature flag `PIPE_DISABLE_PIPELINE_NOTIFICATIONS_ON_ROLLBACK`. [PIPE-28097]
 - Fixed an issue where the CD Git Clone step failed due to an internal mapping error even when valid repository details were provided. [PIPE-28355, ZD-87095]
-- Fixed an issue where the Fargate deployment `EcsUpgradeContainer` step did not perform a health check on retry, resulting in false positives. This is resolved with the feature flag `CDS_ECS_STEADY_STATE_CHECK_ON_RETRIES`. [CDS-111777, ZD-86907]
 - Fixed an issue where the Canary Delete step in Kubernetes deployments deleted Virtual Services during rollback, leading to service downtime. This is resolved with the feature flag `CDS_K8S_NOT_ADD_TRAFFIC_ROUTING_TO_CANARY_WORKLOAD`. [CDS-111916, ZD-87449]
 - Fixed an issue where rollback was triggered even when the last stage (a child pipeline stage) failed with a strategy of Mark as Failure. The rollback is no longer triggered in such cases. [PIPE-28067, ZD-86977]
 - Fixed an issue where a pull request creation from the Pipeline Studio incorrectly showed a failure error message, even though the PR was successfully created. [PIPE-27981]