[PRMP-570] Step Function Error Handling (#486)

Author: SWhyteAnswer

infrastructure/buckets.tf

@@ -85,6 +85,18 @@ module "migration-dynamodb-segment-store" {
   force_destroy             = local.is_force_destroy
 }
 
+module "migration-failed-items-store" {
+  source                    = "./modules/s3/"
+  access_logs_enabled       = local.is_production
+  access_logs_bucket_id     = local.access_logs_bucket_id
+  bucket_name               = var.migration_failed_items_store_bucket_name
+  enable_cors_configuration = false
+  enable_bucket_versioning  = true
+  environment               = var.environment
+  owner                     = var.owner
+  force_destroy             = local.is_force_destroy
+}
+
 module "statistical-reports-store" {
   source                    = "./modules/s3/"
   access_logs_enabled       = local.is_production

infrastructure/lambda-dynamodb-migration.tf

@@ -10,7 +10,8 @@ module "migration-dynamodb-lambda" {
     module.ndr-bulk-staging-store.s3_read_policy_document,
     module.ndr-lloyd-george-store.s3_read_policy_document,
     aws_iam_policy.ssm_access_policy.policy,
-    module.ndr-app-config.app_config_policy
+    module.ndr-app-config.app_config_policy,
+    module.migration-failed-items-store.s3_write_policy_document
   ]
 
   kms_deletion_window           = var.kms_deletion_window
@@ -20,10 +21,11 @@ module "migration-dynamodb-lambda" {
   is_invoked_from_gateway       = false
 
   lambda_environment_variables = {
-    WORKSPACE               = terraform.workspace
-    APPCONFIG_APPLICATION   = module.ndr-app-config.app_config_application_id
-    APPCONFIG_ENVIRONMENT   = module.ndr-app-config.app_config_environment_id
-    APPCONFIG_CONFIGURATION = module.ndr-app-config.app_config_configuration_profile_id
+    WORKSPACE                                = terraform.workspace
+    APPCONFIG_APPLICATION                    = module.ndr-app-config.app_config_application_id
+    APPCONFIG_ENVIRONMENT                    = module.ndr-app-config.app_config_environment_id
+    APPCONFIG_CONFIGURATION                  = module.ndr-app-config.app_config_configuration_profile_id
+    MIGRATION_FAILED_ITEMS_STORE_BUCKET_NAME = "${terraform.workspace}-${var.migration_failed_items_store_bucket_name}"
   }
 
   lambda_timeout                 = 900
@@ -35,5 +37,6 @@ module "migration-dynamodb-lambda" {
     module.bulk_upload_report_dynamodb_table,
     module.ndr-app-config,
     aws_iam_policy.ssm_access_policy,
+    module.migration-failed-items-store
   ]
 }

infrastructure/step-function-migration-dynmodb.tf

@@ -116,7 +116,17 @@ resource "aws_sfn_state_machine" "migration_dynamodb" {
           "key.$"    = "$.Payload.key"
         },
         ResultPath = "$.SegmentSource",
-        Next       = "Segment Map (Distributed)"
+        Next       = "Segment Map (Distributed)",
+
+        Retry = [
+          {
+            ErrorEquals     = ["States.ALL"]
+            IntervalSeconds = 2
+            MaxAttempts     = 3
+            BackoffRate     = 2.0
+            JitterStrategy  = "FULL"
+          }
+        ]
       },
 
       "Segment Map (Distributed)" = {
@@ -139,8 +149,8 @@ resource "aws_sfn_state_machine" "migration_dynamodb" {
           "totalSegments.$"   = "$.totalSegments",
           "tableArn.$"        = "$.tableArn",
           "migrationScript.$" = "$.migrationScript",
-          "run_migration.$"   = "$.run_migration",
-          "execution_Id.$"    = "$$.Execution.Id"
+          "runMigration.$"    = "$.runMigration",
+          "executionId.$"     = "$$.Execution.Id"
         },
 
         ItemProcessor = {
@@ -160,13 +170,22 @@ resource "aws_sfn_state_machine" "migration_dynamodb" {
                   "totalSegments.$"   = "$.totalSegments",
                   "tableArn.$"        = "$.tableArn",
                   "migrationScript.$" = "$.migrationScript",
-                  "run_migration.$"   = "$.run_migration",
-                  "execution_Id.$"    = "$.execution_Id"
+                  "runMigration.$"    = "$.runMigration",
+                  "executionId.$"     = "$.executionId"
                 }
               },
               ResultSelector = { "migrationResult.$" = "$.Payload" },
               ResultPath     = "$.MigrationResult",
-              End            = true
+              End            = true,
+              Retry = [
+                {
+                  ErrorEquals     = ["States.ALL"]
+                  IntervalSeconds = 10
+                  MaxAttempts     = 3
+                  BackoffRate     = 2.0
+                  JitterStrategy  = "FULL"
+                }
+              ]
             }
           }
         },

infrastructure/variable.tf

@@ -39,6 +39,11 @@ variable "migration_dynamodb_segment_store_bucket_name" {
   default     = "migration-dynamodb-segment-store"
 }
 
+variable "migration_failed_items_store_bucket_name" {
+  description = "The name of the S3 bucket to store failed items during DynamoDB migration."
+  type        = string
+  default     = "migration-failed-items-store"
+}
 variable "zip_store_bucket_name" {
   description = "The name of the S3 bucket used as a zip store."
   type        = string