Upgrade to Cumulus 16.1.1, ORCA 8.0.1

Fixes #265

.github/workflows/main.yml

@@ -55,12 +55,12 @@ jobs:
       deploy: true
     secrets: inherit
 
-  deploy-prod:
-    # Run only on main branch (on push, including on merged PRs)
-    if: github.event.ref == 'refs/heads/main'
-    needs: deploy-uat
-    uses: ./.github/workflows/terraspace.yml
-    with:
-      TS_ENV: prod
-      deploy: true
-    secrets: inherit
+  # deploy-prod:
+  #   # Run only on main branch (on push, including on merged PRs)
+  #   if: github.event.ref == 'refs/heads/main'
+  #   needs: deploy-uat
+  #   uses: ./.github/workflows/terraspace.yml
+  #   with:
+  #     TS_ENV: prod
+  #     deploy: true
+  #   secrets: inherit

.nvmrc

@@ -1 +1 @@
-14.19.3
+16.20.2

app/stacks/cumulus/iam.tf

@@ -1,3 +1,21 @@
+locals {
+  # See https://docs.aws.amazon.com/elasticloadbalancing/latest/classic/enable-access-logs.html#attach-bucket-policy
+  # At above link, expand the collapsed section titled "Regions available before
+  # August 2022" to see the list of Elastic Load Balancing account IDs by region.
+  # The following map is constructed from that list, but may not be complete, as
+  # we don't necessarily need to cover all regions.
+  elb_account_ids = {
+    "us-east-1" = "127311923021"
+    "us-east-2" = "033677994240"
+    "us-west-1" = "027434742980"
+    "us-west-2" = "797873946194"
+  }
+  # 797873946194 is the AWS account for Elastic Load Balancing for us-west-2 as
+  # shown in the elb_account_ids map above. This is the default value for the
+  # lookup function used farther below, as a matter of best practice.
+  default_elb_account_id = "797873946194"
+}
+
 # <% if !in_sandbox? then %>
 data "aws_iam_policy_document" "allow_s3_access_logging" {
   statement {
@@ -16,9 +34,15 @@ data "aws_iam_policy_document" "allow_s3_access_logging" {
 }
 # <% end %>
 
+#-------------------------------------------------------------------------------
 # Additional permissions required in order to allow Step Functions to include
-# Distributed Map states.
-# See https://docs.aws.amazon.com/step-functions/latest/dg/use-dist-map-orchestrate-large-scale-parallel-workloads.html#dist-map-permissions
+# Distributed Map states.  This is what allows us to sidestep the 25,000 event-
+# transition quota for Step Functions.
+#
+# See also:
+# - https://docs.aws.amazon.com/step-functions/latest/dg/use-dist-map-orchestrate-large-scale-parallel-workloads.html#dist-map-permissions
+#-------------------------------------------------------------------------------
+
 data "aws_iam_policy_document" "allow_sfn_distributed_maps" {
   statement {
     effect = "Allow"
@@ -50,7 +74,54 @@ resource "aws_iam_role_policy_attachment" "allow_sfn_distributed_maps" {
   policy_arn = aws_iam_policy.allow_sfn_distributed_maps.arn
 }
 
-# temporary workaround for dashboard permissions issue
+#-------------------------------------------------------------------------------
+# Additional policy required on the system bucket as per ORCA v8.0.0.
+#
+# See also:
+# - https://github.com/nasa/cumulus-orca/releases/tag/v8.0.0
+# - https://nasa.github.io/cumulus-orca/docs/developer/deployment-guide/deployment-s3-bucket#bucket-policy-for-load-balancer-server-access-logging
+#-------------------------------------------------------------------------------
+
+data "aws_iam_policy_document" "allow_load_balancer_s3_write_access" {
+  statement {
+    effect  = "Allow"
+    actions = ["s3:PutObject"]
+    principals {
+      type = "AWS"
+      identifiers = [
+        "arn:aws:iam::${lookup(local.elb_account_ids, data.aws_region.current.name, local.default_elb_account_id)}:root"
+      ]
+    }
+    resources = [
+      "arn:aws:s3:::${var.system_bucket}/${var.prefix}-lb-gql-a-logs/AWSLogs/${local.aws_account_id}/*"
+    ]
+  }
+}
+
+# Attach policy above to the system bucket
+resource "null_resource" "allow_load_balancer_s3_write_access" {
+  triggers = {
+    buckets = var.system_bucket
+  }
+
+  # Since we do not have Terraform configured to manage our buckets, we cannot
+  # ask Terraform to put any policies on the buckets, so we're calling out to
+  # the AWS CLI to put the desired policy on our "system" (internal) bucket to
+  # allow load balancer logs to be written to it, as required by ORCA.
+  provisioner "local-exec" {
+    interpreter = ["bash", "-c"]
+    command     = <<-COMMAND
+      aws s3api put-bucket-policy \
+        --bucket ${var.system_bucket} \
+        --policy '${data.aws_iam_policy_document.allow_load_balancer_s3_write_access.json}'
+    COMMAND
+  }
+}
+
+#-------------------------------------------------------------------------------
+# Temporary workaround for dashboard permissions issue
+#-------------------------------------------------------------------------------
+
 data "aws_iam_role" "api_gateway_role" {
   depends_on = [module.cumulus]
 

app/stacks/cumulus/main.tf

@@ -483,7 +483,7 @@ module "ingest_and_publish_granule_workflow" {
     files_to_granules_task_arn : module.cumulus.files_to_granules_task.task_arn,
     move_granules_task_arn : module.cumulus.move_granules_task.task_arn,
     update_granules_cmr_metadata_file_links_task_arn : module.cumulus.update_granules_cmr_metadata_file_links_task.task_arn,
-    copy_to_archive_task_arn : module.orca.orca_lambda_copy_to_glacier_arn,
+    copy_to_archive_adapter_task_arn : module.cumulus.orca_copy_to_archive_adapter_task.task_arn,
     post_to_cmr_task_arn : module.cumulus.post_to_cmr_task.task_arn
   })
 }
@@ -495,6 +495,7 @@ module "cumulus" {
   deploy_to_ngap = true
 
   cumulus_message_adapter_lambda_layer_version_arn = module.cma.lambda_layer_version_arn
+  async_operation_image                            = "cumuluss/async-operation:47"
 
   vpc_id            = module.vpc.vpc_id
   lambda_subnet_ids = module.vpc.subnets.ids
@@ -508,6 +509,10 @@ module "cumulus" {
   ecs_cluster_max_size            = 2
   key_name                        = var.key_name
 
+  orca_api_uri                    = module.orca.orca_api_deployment_invoke_url
+  orca_lambda_copy_to_archive_arn = module.orca.orca_lambda_copy_to_archive_arn
+  orca_sfn_recovery_workflow_arn  = module.orca.orca_sfn_recovery_workflow_arn
+
   rds_security_group         = local.rds_security_group
   rds_user_access_secret_arn = local.rds_user_access_secret_arn
 

app/stacks/cumulus/orca.tf

@@ -15,7 +15,7 @@ data "aws_secretsmanager_secret_version" "rds_cluster_user_credentials_secret_ve
 }
 
 module "orca" {
-  source = "https://github.com/nasa/cumulus-orca/releases/download/v6.0.3/cumulus-orca-terraform.zip"
+  source = "https://github.com/nasa/cumulus-orca/releases/download/v8.0.1/cumulus-orca-terraform.zip"
   #--------------------------
   # Cumulus variables
   #--------------------------
@@ -53,7 +53,7 @@ module "orca" {
   # default_multipart_chunksize_mb                       = 250
   # metadata_queue_message_retention_time                = 777600
   # orca_default_recovery_type                           = "Standard"
-  orca_default_storage_class                           = "DEEP_ARCHIVE"
+  orca_default_storage_class = "DEEP_ARCHIVE"
   # orca_delete_old_reconcile_jobs_frequency_cron        = "cron(0 0 ? * SUN *)"
   # orca_ingest_lambda_memory_size                       = 2240
   # orca_ingest_lambda_timeout                           = 600

app/stacks/cumulus/templates/ingest-and-publish-granule-workflow.asl.json

@@ -233,8 +233,8 @@
                 "cma": {
                   "event.$": "$",
                   "task_config": {
-                    "excludedFileExtensions": "{$.meta.collection.meta.orca.excludedFileExtensions}",
                     "s3MultipartChunksizeMb": "{$.meta.collection.meta.s3MultipartChunksizeMb}",
+                    "excludedFileExtensions": "{$.meta.collection.meta.orca.excludedFileExtensions}",
                     "providerId": "{$.meta.provider.id}",
                     "providerName": "{$.meta.provider.name}",
                     "executionId": "{$.cumulus_meta.execution_name}",
@@ -245,7 +245,7 @@
                 }
               },
               "Type": "Task",
-              "Resource": "${copy_to_archive_task_arn}",
+              "Resource": "${copy_to_archive_adapter_task_arn}",
               "Retry": [
                 {
                   "ErrorEquals": [

app/stacks/rds-cluster/main.tf

@@ -4,6 +4,14 @@ resource "random_password" "db_password" {
   special = false
 }
 
+resource "random_password" "db_user_password" {
+  length  = 50
+  upper   = true
+  lower   = true
+  number  = true
+  special = false
+}
+
 module "vpc" {
   source = "../../modules/vpc"
 }
@@ -15,15 +23,21 @@ module "rds_cluster" {
   db_admin_password        = random_password.db_password.result
   db_admin_username        = "postgres"
   deletion_protection      = true
-  engine_version           = "11.13"
+  engine_version           = "11.18"
   parameter_group_family   = "aurora-postgresql11"
   permissions_boundary_arn = local.permissions_boundary_arn
   prefix                   = var.prefix
   provision_user_database  = true
-  rds_user_password        = random_password.db_password.result
-  region                   = data.aws_region.current.name
-  snapshot_identifier      = null
-  subnets                  = module.vpc.subnets.ids
-  tags                     = { Deployment = var.prefix }
-  vpc_id                   = module.vpc.vpc_id
+  # ORCA requires us to use a password that contains a special character, but there is
+  # some Cumulus constraint that only allows an underscore (in addition to alphanumeric
+  # characters), and no other special characters, so we must generate a password that
+  # does not contain any special characters, in order to avoid special characters other
+  # than and underscore, and then insert an underscore (we chose to at it to the end) to
+  # satisfy the ORCA constraint requiring at least one special character.
+  rds_user_password   = "${random_password.db_user_password.result}_"
+  region              = data.aws_region.current.name
+  snapshot_identifier = null
+  subnets             = module.vpc.subnets.ids
+  tags                = { Deployment = var.prefix }
+  vpc_id              = module.vpc.vpc_id
 }

config/helpers/cumulus_version_helper.rb

@@ -1,5 +1,5 @@
 module Terraspace::Project::CumulusVersionHelper
   def cumulus_version
-    "v13.4.0"
+    "v16.1.1"
   end
 end

package.json

@@ -52,30 +52,35 @@
     "tf:lambda:archive-exploded": "yarn lambda:archive-exploded >&2 && echo { '\"'dir'\"': '\"'${PWD}/${npm_package_config_lambda_archive_dir}'\"' }"
   },
   "engines": {
-    "node": ">=14"
+    "node": ">=16"
   },
   "dependencies": {
     "@aws-sdk/client-s3": "^3.370.0",
-    "@cumulus/aws-client": "13.4.0",
-    "@cumulus/cmrjs": "13.4.0",
-    "@cumulus/common": "13.4.0",
+    "@aws-sdk/client-sts": "^3.370.0",
+    "@aws-sdk/lib-dynamodb": "^3.370.0",
+    "@aws-sdk/lib-storage": "^3.370.0",
+    "@aws-sdk/types": "^3.370.0",
+    "@cumulus/aws-client": "16.1.1",
+    "@cumulus/cmrjs": "16.1.1",
+    "@cumulus/common": "16.1.1",
     "@cumulus/cumulus-message-adapter-js": "2.0.4",
+    "@smithy/util-stream": "^2.0.17",
     "date-fns": "^2.29.3",
     "duration-fns": "^3.0.1",
     "fp-ts": "^2.11.5",
     "fp-ts-contrib": "^0.1.29",
     "io-ts": "^2.2.16",
     "io-ts-types": "^0.5.16",
     "monocle-ts": "^2.3.13",
-    "newtype-ts": "^0.3.5"
+    "newtype-ts": "^0.3.5",
+    "xml2js": "^0.6.0"
   },
   "devDependencies": {
     "@ava/typescript": "^3.0.1",
     "@aws-sdk/client-dynamodb": "^3.370.0",
-    "@aws-sdk/util-stream-node": "^3.370.0",
-    "@cumulus/types": "13.4.0",
+    "@cumulus/types": "16.1.1",
     "@istanbuljs/nyc-config-typescript": "^1.0.1",
-    "@tsconfig/node14": "^1.0.1",
+    "@tsconfig/node16": "^16.1.1",
     "@types/aws-lambda": "^8.10.85",
     "@types/lodash": "^4.14.177",
     "@types/node": "^16.11.1",
@@ -97,7 +102,7 @@
     "source-map-support": "^0.5.19",
     "standard-version": "^9.0.0",
     "ts-node": "^10.3.0",
-    "typedoc": "^0.22.6",
+    "typedoc": "^0.25.3",
     "typescript": "^4.4.4"
   },
   "resolutions": {

src/lib/aws/s3.fixture.ts

@@ -1,7 +1,7 @@
 import { Readable } from 'stream';
 
 import { GetObjectCommandInput } from '@aws-sdk/client-s3';
-import { sdkStreamMixin } from '@aws-sdk/util-stream-node';
+import { sdkStreamMixin } from '@smithy/util-stream';
 
 const store: { readonly [Bucket: string]: { readonly [Key: string]: string } } = {
   'my-bucket': {