Merge pull request #2700 from nick-stroud/ofe_new_features_alt_history

OFE: new features - edited history

Author: nick-stroud

community/front-end/ofe/deploy.sh

@@ -57,6 +57,8 @@ PRJ_API['bigqueryconnection.googleapis.com']='BigQuery Connection API'
 PRJ_API['sqladmin.googleapis.com']='Cloud SQL Admin API'
 PRJ_API['servicenetworking.googleapis.com']='Service Networking API'
 PRJ_API['secretmanager.googleapis.com']='Secret Manager API'
+PRJ_API['serviceusage.googleapis.com']='Service Usage API'
+PRJ_API['storage.googleapis.com']='Cloud Storage API'
 
 # Location for output credential file = pwd/credential.json
 #

community/front-end/ofe/infrastructure_files/gcs_bucket/webserver/startup.sh

@@ -249,6 +249,7 @@ autostart=true
 autorestart=true
 user=gcluster
 redirect_stderr=true
+environment=HOME=/opt/gcluster
 stdout_logfile=/opt/gcluster/run/supvisor.log" >/etc/supervisord.d/gcluster.ini
 
 printf "Creating systemd service..."

community/front-end/ofe/infrastructure_files/vpc_tf/GCP/README.md

@@ -19,13 +19,15 @@ limitations under the License.
 |------|---------|
 | <a name="requirement_terraform"></a> [terraform](#requirement\_terraform) | >= 0.12.31 |
 | <a name="requirement_google"></a> [google](#requirement\_google) | >= 3.54 |
+| <a name="requirement_google-beta"></a> [google-beta](#requirement\_google-beta) | >= 3.83 |
 | <a name="requirement_random"></a> [random](#requirement\_random) | >= 3.0 |
 
 ## Providers
 
 | Name | Version |
 |------|---------|
 | <a name="provider_google"></a> [google](#provider\_google) | >= 3.54 |
+| <a name="provider_google-beta"></a> [google-beta](#provider\_google-beta) | >= 3.83 |
 | <a name="provider_random"></a> [random](#provider\_random) | >= 3.0 |
 
 ## Modules
@@ -36,11 +38,14 @@ No modules.
 
 | Name | Type |
 |------|------|
+| [google-beta_google_compute_global_address.private_ip_alloc](https://registry.terraform.io/providers/hashicorp/google-beta/latest/docs/resources/google_compute_global_address) | resource |
 | [google_compute_firewall.firewall_allow_ssh](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_firewall) | resource |
 | [google_compute_firewall.firewall_internal](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_firewall) | resource |
 | [google_compute_network.network](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_network) | resource |
 | [google_compute_router.network_router](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_router) | resource |
 | [google_compute_router_nat.network_nat](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/compute_router_nat) | resource |
+| [google_service_networking_connection.private_vpc_connection](https://registry.terraform.io/providers/hashicorp/google/latest/docs/resources/service_networking_connection) | resource |
+| [random_id.resource_name_suffix](https://registry.terraform.io/providers/hashicorp/random/latest/docs/resources/id) | resource |
 | [random_pet.vpc_name](https://registry.terraform.io/providers/hashicorp/random/latest/docs/resources/pet) | resource |
 
 ## Inputs

community/front-end/ofe/infrastructure_files/vpc_tf/GCP/main.tf

@@ -85,6 +85,34 @@ resource "google_compute_firewall" "firewall_internal" {
   allow { protocol = "icmp" }
 }
 
+locals {
+  # This label allows for billing report tracking based on module.
+  labels = {
+    created_by = "ofe"
+  }
+}
+
+resource "random_id" "resource_name_suffix" {
+  byte_length = 4
+}
+
+resource "google_compute_global_address" "private_ip_alloc" {
+  provider      = google-beta
+  project       = var.project
+  name          = "global-psconnect-ip-${random_id.resource_name_suffix.hex}"
+  purpose       = "VPC_PEERING"
+  address_type  = "INTERNAL"
+  network       = google_compute_network.network.self_link
+  prefix_length = 16
+  labels        = local.labels
+}
+
+resource "google_service_networking_connection" "private_vpc_connection" {
+  network                 = google_compute_network.network.self_link
+  service                 = "servicenetworking.googleapis.com"
+  reserved_peering_ranges = [google_compute_global_address.private_ip_alloc.name]
+}
+
 output "vpc_id" {
   value       = google_compute_network.network.name
   description = "Name of the created VPC"

community/front-end/ofe/infrastructure_files/vpc_tf/GCP/versions.tf

@@ -20,6 +20,10 @@ terraform {
       source  = "hashicorp/google"
       version = ">= 3.54"
     }
+    google-beta = {
+      source  = "hashicorp/google-beta"
+      version = ">= 3.83"
+    }
     random = {
       source  = "hashicorp/random"
       version = ">= 3.0"

community/front-end/ofe/website/ghpcfe/cluster_manager/cloud_info.py

@@ -39,6 +39,7 @@
         "n1": defaultdict(lambda: "x86_64"),
         "c3": defaultdict(lambda: "sapphirerapids"),
         "c3d": defaultdict(lambda: "zen2"),
+        "c4": defaultdict(lambda: "emeraldrapids"),
         # Compute Optimized
         "c2": defaultdict(lambda: "cascadelake"),
         "c2d": defaultdict(
@@ -359,6 +360,7 @@ def get_cpu_price(num_cores, instance_type, skus):
             "n2d": "N2D AMD Instance Core",
             "h3": "Compute optimized Core",
             "c3": "Compute optimized Core",
+            "c4": "Compute optimized Core",
             "c2": "Compute optimized Core",
             "c2d": "C2D AMD Instance Core",
             "c3d": "C3D AMD Instance Core",
@@ -411,6 +413,7 @@ def get_mem_price(num_gb, instance_type, skus):
             "h3": "Compute optimized Ram",
             "c2d": "C2D AMD Instance Ram",
             "c3d": "C3D AMD Instance Ram",
+            "c4": "C4 Instance RAM",
             "t2d": "T2D AMD Instance Ram",
             "a2": "A2 Instance Ram",
             "m1": "Memory-optimized Instance Ram",

community/front-end/ofe/website/ghpcfe/forms.py

@@ -247,8 +247,10 @@ class Meta:
             "image",
             "dynamic_node_count",
             "static_node_count",
+            "reservation_name",
             "enable_placement",
             "enable_hyperthreads",
+            "enable_tier1_networking",
             "enable_node_reuse",
             "GPU_type",
             "GPU_per_node",
@@ -303,6 +305,9 @@ def prep_dynamic_select(field, value):
             self.instance.GPU_type
         )
 
+        # Mark 'reservation_name' as optional
+        self.fields["reservation_name"].widget.attrs.update({"placeholder": "Optional"})
+
     def clean(self):
         cleaned_data = super().clean()
         if cleaned_data["enable_placement"] and cleaned_data[

community/front-end/ofe/website/ghpcfe/models.py

@@ -739,9 +739,9 @@ class Cluster(CloudResource):
         default="pd-standard",
     )
     controller_disk_size = models.PositiveIntegerField(
-        validators=[MinValueValidator(10)],
+        validators=[MinValueValidator(120)],
         help_text="Boot disk size (in GB)",
-        default=50,
+        default=120,
         blank=True,
     )
     num_login_nodes = models.PositiveIntegerField(
@@ -762,9 +762,9 @@ class Cluster(CloudResource):
     login_node_disk_size = models.PositiveIntegerField(
         # login node disk must be large enough to hold the SlurmGCP
         # image: >=50GB
-        validators=[MinValueValidator(50)],
+        validators=[MinValueValidator(120)],
         help_text="Boot disk size (in GB)",
-        default=50,
+        default=120,
         blank=True,
     )
     grafana_dashboard_url = models.CharField(
@@ -919,6 +919,13 @@ class ClusterPartition(models.Model):
     enable_hyperthreads = models.BooleanField(
         default=False, help_text="Enable Hyperthreads (SMT)"
     )
+    enable_tier1_networking = models.BooleanField(
+        default=False,
+        help_text=(
+            "Select Tier 1 networking (currently only valid for N2, N2D, C2,"
+            "C2D, C3, C3d, M3 and Z3 VMs that have at least 30 vCPUs.)"
+        ),
+    )
     enable_node_reuse = models.BooleanField(
         default=True,
         help_text=(
@@ -937,7 +944,7 @@ class ClusterPartition(models.Model):
         default="pd-standard",
     )
     boot_disk_size = models.PositiveIntegerField(
-        validators=[MinValueValidator(49)],
+        validators=[MinValueValidator(50)],
         help_text="Boot disk size (in GB)",
         default=50,
         blank=True,
@@ -972,7 +979,11 @@ class ClusterPartition(models.Model):
             "Automatically delete additional disk when node is deleted?"
         ),
     )
-
+    reservation_name = models.CharField(
+        blank=True,
+        max_length=30,
+        help_text="Name of the reservation to use for VM resources"
+    )
     def __str__(self):
         return self.name
 
@@ -1572,9 +1583,9 @@ class Workbench(CloudResource):
         help_text="Type of storage to be required for notebook boot disk",
     )
     boot_disk_capacity = models.PositiveIntegerField(
-        validators=[MinValueValidator(100)],
+        validators=[MinValueValidator(120)],
         help_text="Capacity (in GB) of the filesystem (min of 1024)",
-        default=100,
+        default=120,
     )
     proxy_uri = models.CharField(max_length=150, blank=True, null=True)
     trusted_user = models.ForeignKey(

community/front-end/ofe/website/ghpcfe/templates/blueprint/cloudsql_config.yaml.j2

@@ -1,7 +1,7 @@
 - source: community/modules/database/slurm-cloudsql-federation
   kind: terraform
   id: slurm-sql
-  use: [hpc_network, ps-connect]
+  use: [hpc_network]
   settings:
     sql_instance_name: sql-{{ cluster_id }}
     tier: "db-g1-small"

community/front-end/ofe/website/ghpcfe/templates/blueprint/cluster_config.yaml.j2

@@ -23,10 +23,6 @@ deployment_groups:
       subnetwork_name: {{ cluster.subnet.cloud_id }}
     id: hpc_network
 
-  - source: community/modules/network/private-service-access
-    id: ps-connect
-    use: [ hpc_network ]    
-
 {{ filesystems_yaml | safe }}
 
   - source: community/modules/project/service-account

community/front-end/ofe/website/ghpcfe/templates/blueprint/partition_config.yaml.j2

@@ -14,8 +14,12 @@
   id: {{ part_id }}-group
   use:
   settings:
+    bandwidth_tier: {% if part.enable_tier1_networking %}tier_1_enabled{% else %}platform_default{% endif %}
     enable_smt: {{ part.enable_hyperthreads }}
     machine_type: {{ part.machine_type }}
+    {% if part.reservation_name %}
+    reservation_name: {{ part.reservation_name }}
+    {% endif %}
     node_count_dynamic_max: {{ part.dynamic_node_count }}
     node_count_static: {{ part.static_node_count }}
     disk_size_gb: {{ part.boot_disk_size }}