Increment arthur-engine version (#904)

Author: alexshanabrook

.gitignore

@@ -38,3 +38,5 @@ values.yaml
 
 ml-engine/src/genai_client/**
 .vagrant
+
+.cursor/

deployment/README_RELEASE.md

@@ -0,0 +1,18 @@
+# Release Process
+
+## Prerequisites
+- Ensure all necessary commits are merged to the `dev` branch
+- Verify integration tests are passing for the ML engine by running them locally
+- Verify integration tests are passing for the GenAI engine by running them locally
+
+## Dev Release
+1. In GitHub Actions, manually run the **Version Management** workflow with the default inputs (`dev` branch, patch bump)
+2. This will create a dev release and trigger the Arthur Engine CI workflow
+
+## Production Release
+1. After the dev release pipeline has succeeded, create a PR to merge `dev` into `main`
+2. **Important**: The version bump commit must be the most recent commit in this PR, or the release won't be triggered
+3. Merge the PR to trigger the production release
+4. Let Madeleine know to publish the release notes on GitHub.
+
+> **Note**: The version bump commit requirement prevents unintended commits from being included in the release.

deployment/cloudformation/root-arthur-engine-cpu.yml

@@ -319,6 +319,14 @@ Parameters:
     Type: Number
     Description: 'The desired number of tasks running for ML Engine ECS service'
     Default: 2
+  MLEngineFargateTaskCPUValue:
+    Type: Number
+    Default: 2048
+    Description: 'CPU value for AWS Fargate per ML Engine ECS task'
+  MLEngineFargateTaskMemoryValue:
+    Type: Number
+    Default: 4096
+    Description: 'Memory value for AWS Fargate per ML Engine ECS task'
 Metadata:
   AWS::CloudFormation::Interface:
     ParameterGroups:
@@ -404,10 +412,10 @@ Metadata:
           default: 'Advanced ML Engine'
         Parameters:
           - MLEngineVersion
-          - MLEngineClientId
-          - MLEngineClientSecret
           - MLEngineContainerImageLocation
           - MLEngineBYOAppSecurityGroupIDs
+          - MLEngineFargateTaskCPUValue
+          - MLEngineFargateTaskMemoryValue
       - Label:
           default: 'Advanced ML Engine (IAM)'
         Parameters:
@@ -563,6 +571,10 @@ Metadata:
         default: 'Bring-Your-Own ML Engine Task Execution Role'
       MLEngineBYOAppSecurityGroupIDs:
         default: 'Bring-Your-Own ML Engine Application Security Group IDs'
+      MLEngineFargateTaskCPUValue:
+        default: 'ML Engine ECS CPU Value. Must be compatible with ECS Memory Value'
+      MLEngineFargateTaskMemoryValue:
+        default: 'ML Engine ECS Memory Value. Must be compatible with ECS CPU Value'
       ArthurApiHost:
         default: 'Arthur Platform URL'
 Conditions:
@@ -743,6 +755,8 @@ Resources:
         MLEngineClientCredentialsSecretARN: !GetAtt [ MLEngineSecretsStack, Outputs.MLEngineClientCredentialsSecretOutput ]
         GenaiEngineInternalAPIKeySecretARN: !GetAtt [ GenaiEngineSecretsStack, Outputs.GenaiEngineAPIKeySecretOutput ]
         GenaiEngineInternalIngressHost: !Ref GenaiEngineIngressURL
+        MLEngineFargateTaskCPUValue: !Ref MLEngineFargateTaskCPUValue
+        MLEngineFargateTaskMemoryValue: !Ref MLEngineFargateTaskMemoryValue
       TemplateURL: "https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/REPLACE_ME_GENAI_ENGINE_VERSION/ml-engine/arthur-ml-engine-ecs-task-definition.yml"
   GenaiEngineLBStack:
     Type: AWS::CloudFormation::Stack

deployment/cloudformation/root-arthur-engine-gpu.yml

@@ -433,6 +433,14 @@ Parameters:
     Type: Number
     Description: 'The desired number of tasks running for ML Engine ECS service'
     Default: 2
+  MLEngineFargateTaskCPUValue:
+    Type: Number
+    Default: 2048
+    Description: 'CPU value for AWS Fargate per ML Engine ECS task'
+  MLEngineFargateTaskMemoryValue:
+    Type: Number
+    Default: 4096
+    Description: 'Memory value for AWS Fargate per ML Engine ECS task'
 Metadata:
   AWS::CloudFormation::Interface:
     ParameterGroups:
@@ -540,13 +548,39 @@ Metadata:
           - GenaiEngineHallucinationCheckMaxTokenLimit
           - GenaiEngineToxicityCheckMaxTokenLimit
           - GenaiEngineUsePIIModelV2
+      - Label:
+          default: 'Advanced ML Engine'
+        Parameters:
+          - MLEngineVersion
+          - MLEngineContainerImageLocation
+          - MLEngineBYOAppSecurityGroupIDs
+          - MLEngineFargateTaskCPUValue
+          - MLEngineFargateTaskMemoryValue
+      - Label:
+          default: 'Advanced ML Engine (IAM)'
+        Parameters:
+          - MLEngineBYOTaskRoleIAMArn
+          - MLEngineBYOTaskExecutionRoleIAMArn
+      - Label:
+          default: 'Advanced ML Engine (Scalability)'
+        Parameters:
+          - MLEngineECSServiceTaskDesiredCount
+          - MLEngineECSAutoscalingMinCapacity
+          - MLEngineECSAutoscalingMaxCapacity
+          - MLEngineECSScaleCPUTargetValue
+          - MLEngineECSScaleMemoryTargetValue
+          - MLEngineECSScaleInCooldownInSecs
+          - MLEngineECSScaleOutCooldownInSecs
+      - Label:
+          default: 'Advanced ML Engine (Platform Monitoring)'
+        Parameters:
+          - MLEngineECSCPUAlarmThreshold
+          - MLEngineECSMemoryAlarmThreshold
     ParameterLabels:
       ArthurResourceNamespace:
         default: 'Arthur Platform Resource Namespace'
       ArthurResourceNameSuffix:
         default: 'Arthur Platform Resource Name Suffix'
-      MLEngineContainerImageLocation:
-        default: 'ML Engine Container Image Location'
       ContainerRepositoryUsername:
         default: 'Container Repository Username'
       ContainerRepositoryPassword:
@@ -723,6 +757,10 @@ Metadata:
         default: 'ML Engine ECS Autoscaling Minimum Capacity'
       MLEngineECSServiceTaskDesiredCount:
         default: 'ML Engine ECS Service Task Desired Count'
+      MLEngineFargateTaskCPUValue:
+        default: 'ML Engine ECS CPU Value. Must be compatible with ECS Memory Value'
+      MLEngineFargateTaskMemoryValue:
+        default: 'ML Engine ECS Memory Value. Must be compatible with ECS CPU Value'
 Conditions:
   ContainerRepositoryCredentialRequired:
     !Not [ !Equals [ !Ref ContainerRepositoryUsername, '' ] ]
@@ -900,6 +938,8 @@ Resources:
         MLEngineClientCredentialsSecretARN: !GetAtt [ MLEngineSecretsStack, Outputs.MLEngineClientCredentialsSecretOutput ]
         GenaiEngineInternalAPIKeySecretARN: !GetAtt [ GenaiEngineSecretsStack, Outputs.GenaiEngineAPIKeySecretOutput ]
         GenaiEngineInternalIngressHost: !Ref GenaiEngineIngressURL
+        MLEngineFargateTaskCPUValue: !Ref MLEngineFargateTaskCPUValue
+        MLEngineFargateTaskMemoryValue: !Ref MLEngineFargateTaskMemoryValue
       TemplateURL: "https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/REPLACE_ME_GENAI_ENGINE_VERSION/ml-engine/arthur-ml-engine-ecs-task-definition.yml"
   GenaiEngineLBStack:
     Type: AWS::CloudFormation::Stack

deployment/cloudformation/start-cpu-stack.html

@@ -1,3 +1,3 @@
-<a href="https://console.aws.amazon.com/cloudformation/home#/stacks/create/review?templateURL=https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/2.1.209/root-arthur-engine-cpu.yml&stackName=arthur-engine&param_MLEngineClientId=yourClientIdHere">
+<a href="https://console.aws.amazon.com/cloudformation/home#/stacks/create/review?templateURL=https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/2.1.235/root-arthur-engine-cpu.yml&stackName=arthur-engine&param_MLEngineClientId=yourClientIdHere">
     <img src="https://s3.amazonaws.com/cloudformation-examples/cloudformation-launch-stack.png" alt="Launch">
 </a>

deployment/cloudformation/start-gpu-stack.html

@@ -1,3 +1,3 @@
-<a href="https://console.aws.amazon.com/cloudformation/home#/stacks/create/review?templateURL=https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/2.1.209/root-arthur-engine-gpu.yml&stackName=arthur-engine&param_MLEngineClientId=yourClientIdHere">
+<a href="https://console.aws.amazon.com/cloudformation/home#/stacks/create/review?templateURL=https://arthur-cft.s3.us-east-2.amazonaws.com/arthur-engine/templates/2.1.235/root-arthur-engine-gpu.yml&stackName=arthur-engine&param_MLEngineClientId=yourClientIdHere">
     <img src="https://s3.amazonaws.com/cloudformation-examples/cloudformation-launch-stack.png" alt="Launch">
 </a>

deployment/helm/arthur-engine/Chart.yaml

@@ -1,11 +1,11 @@
 # Chart.yaml
 apiVersion: v2
 name: arthur-engine
-version: 2.1.209
+version: 2.1.235
 dependencies:
   - name: arthur-genai-engine
-    version: 2.1.209
+    version: 2.1.235
     repository: oci://ghcr.io/arthur-ai/arthur-engine/charts
   - name: arthur-ml-engine
-    version: 2.1.209
+    version: 2.1.235
     repository: oci://ghcr.io/arthur-ai/arthur-engine/charts

deployment/helm/arthur-engine/start.sh.template.cpu

@@ -1,7 +1,7 @@
 #!/bin/bash
 
 K8S_NAMESPACE=arthur
-ARTHUR_ENGINE_VERSION=2.1.209
+ARTHUR_ENGINE_VERSION=2.1.235
 POSTGRES_USER=arthur_genai_engine
 POSTGRES_PASSWORD=changeme_pg_password
 POSTGRES_ENDPOINT=mydb

deployment/helm/arthur-engine/start.sh.template.gpu

@@ -1,7 +1,7 @@
 #!/bin/bash
 
 K8S_NAMESPACE=arthur
-ARTHUR_ENGINE_VERSION=2.1.209
+ARTHUR_ENGINE_VERSION=2.1.235
 POSTGRES_USER=arthur_genai_engine
 POSTGRES_PASSWORD=changeme_pg_password
 POSTGRES_ENDPOINT=mydb

deployment/helm/arthur-engine/values.yaml.template

@@ -96,7 +96,7 @@ arthur-genai-engine:
   ## Advanced - GenAI Engine
   ###################################
   # GenAI Engine version
-  genaiEngineVersion: 2.1.209
+  genaiEngineVersion: 2.1.235
   # Max number of tokens GenAI Engine can process against LLM for hallucination data checks
   genaiEngineHallucinationCheckMaxTokenLimit: 3000
   # Max number of tokens GenAI Engine can process against LLM for sensitive data checks