Version 1.10.0

.gitignore

@@ -7,6 +7,11 @@ ghpc
 # workspace level vscode settings
 .vscode/
 
+####
+community/front-end/ofe/credential.json
+**/.tkfe.lock
+**/deployment.tar.gz
+
 #### TERRAFORM
 
 # Local .terraform directories
@@ -25,6 +30,7 @@ crash.log
 # to change depending on the environment.
 #
 *.tfvars
+**/tf*.log
 
 # Ignore override files as they are usually used to override modules locally and so
 # are not checked in

cmd/create.go

@@ -23,6 +23,7 @@ import (
 	"hpc-toolkit/pkg/config"
 	"hpc-toolkit/pkg/modulewriter"
 	"log"
+	"os"
 
 	"github.com/spf13/cobra"
 )
@@ -98,6 +99,7 @@ func runCreateCmd(cmd *cobra.Command, args []string) {
 		var target *modulewriter.OverwriteDeniedError
 		if errors.As(err, &target) {
 			fmt.Printf("\n%s\n", err.Error())
+			os.Exit(1)
 		} else {
 			log.Fatal(err)
 		}

cmd/root.go

@@ -49,7 +49,7 @@ HPC deployments on the Google Cloud Platform.`,
 				log.Fatalf("cmd.Help function failed: %s", err)
 			}
 		},
-		Version:     "v1.9.0",
+		Version:     "v1.10.0",
 		Annotations: annotation,
 	}
 )

community/examples/README.md

@@ -47,6 +47,10 @@ Examples using Intel HPC technologies can be found in the
 
 [See description in core](../../examples/README.md#slurm-gcp-v5-ubuntu2004yaml-)
 
+### slurm-gcp-v5-high-io.yaml
+
+[See description in core](../../examples/README.md#slurm-gcp-v5-high-ioyaml-)
+
 ### htcondor-pool.yaml
 
 [See description in core](../../examples/README.md#htcondor-poolyaml--)

community/examples/omnia-cluster.yaml

@@ -91,3 +91,4 @@ deployment_groups:
     source: community/modules/scripts/wait-for-startup
     settings:
       instance_name: ((module.manager.name[0]))
+      timeout: 2100

community/examples/slurm-gcp-v5-high-io.yaml

@@ -0,0 +1,150 @@
+# Copyright 2022 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+---
+
+blueprint_name: hpc-cluster-high-io-v5
+
+vars:
+  project_id:  ## Set GCP Project ID Here ##
+  deployment_name: high-io-slurm-gcp-v5
+  region: us-central1
+  zone: us-central1-c
+  # By default, public IPs are set in the login and controller to allow easier
+  # SSH access. To turn this behavior off, set this to true.
+  disable_public_ips: false
+  # Set to true for active cluster reconfiguration. Note that setting this
+  # option requires additional dependencies to be installed locally.
+  enable_reconfigure: true
+
+# Documentation for each of the modules used below can be found at
+# https://github.com/GoogleCloudPlatform/hpc-toolkit/blob/main/modules/README.md
+
+deployment_groups:
+- group: primary
+  modules:
+  # Source is an embedded module, denoted by "modules/*" without ./, ../, /
+  # as a prefix. To refer to a local or community module, prefix with ./, ../ or /
+  # Example - ./modules/network/pre-existing-vpc
+  - id: network1
+    source: modules/network/vpc
+
+  - id: homefs
+    source: modules/file-system/filestore
+    use: [network1]
+    settings:
+      local_mount: /home
+
+  - id: projectsfs
+    source: modules/file-system/filestore
+    use: [network1]
+    settings:
+      filestore_tier: HIGH_SCALE_SSD
+      size_gb: 10240
+      local_mount: /projects
+
+  - id: scratchfs
+    source: community/modules/file-system/DDN-EXAScaler
+    use: [network1]
+    settings:
+      local_mount: /scratch
+
+  # The lowcost partition is designed to run at a lower cost and without additional quota
+  # Use:
+  # `srun -N 4 <<Command>>` for any node in the partition.
+  # `srun -N 4 --mincpus 2` for node group n2s4.
+  - id: low_cost_node_group_n2s2
+    source: community/modules/compute/schedmd-slurm-gcp-v5-node-group
+    settings:
+      name: n2s2
+      machine_type: n2-standard-2
+      node_count_dynamic_max: 10
+
+  - id: low_cost_node_group_n2s4
+    source: community/modules/compute/schedmd-slurm-gcp-v5-node-group
+    settings:
+      name: n2s4
+      machine_type: n2-standard-4
+      node_count_dynamic_max: 10
+
+  - id: low_cost_partition
+    source: community/modules/compute/schedmd-slurm-gcp-v5-partition
+    use:
+    - network1
+    - homefs
+    - scratchfs
+    - projectsfs
+    - low_cost_node_group_n2s2
+    - low_cost_node_group_n2s4
+    settings:
+      is_default: true
+      partition_name: lowcost
+      enable_placement: false
+      exclusive: false
+
+  # The compute partition is designed for performance.
+  # Use:
+  # `srun -N 4 -p compute <<Command>>` for any node in the partition.
+  # `srun -N 4 -p compute --mincpus 30 <<Command>>` for node group c2s60.
+
+  - id: compute_node_group_c2s60
+    source: community/modules/compute/schedmd-slurm-gcp-v5-node-group
+    settings:
+      name: c2s60
+      node_count_dynamic_max: 200
+
+  - id: compute_node_group_c2s30
+    source: community/modules/compute/schedmd-slurm-gcp-v5-node-group
+    settings:
+      name: c2s30
+      node_count_dynamic_max: 200
+      machine_type: c2-standard-30
+
+  - id: compute_partition
+    source: community/modules/compute/schedmd-slurm-gcp-v5-partition
+    use:
+    - network1
+    - homefs
+    - scratchfs
+    - projectsfs
+    - compute_node_group_c2s60
+    - compute_node_group_c2s30
+    settings:
+      partition_name: compute
+
+  - id: slurm_controller
+    source: community/modules/scheduler/schedmd-slurm-gcp-v5-controller
+    use:
+    - network1
+    - homefs
+    - scratchfs
+    - projectsfs
+    - low_cost_partition
+    - compute_partition
+    settings:
+      machine_type: c2-standard-8
+      disable_controller_public_ips: $(vars.disable_public_ips)
+
+  - id: slurm_login
+    source: community/modules/scheduler/schedmd-slurm-gcp-v5-login
+    use:
+    - network1
+    - slurm_controller
+    settings:
+      machine_type: n2-standard-4
+      disable_login_public_ips: $(vars.disable_public_ips)
+
+  - id: hpc_dashboard
+    source: modules/monitoring/dashboard
+    outputs: [instructions]

community/examples/spack-gromacs.yaml

@@ -49,8 +49,6 @@ deployment_groups:
     source: community/modules/scripts/spack-install
     settings:
       install_dir: /sw/spack
-      spack_url: https://github.com/spack/spack
-      spack_ref: v0.18.0
       log_file: /var/log/spack.log
       configs:
       - type: single-config

community/front-end/.gitignore

@@ -0,0 +1,27 @@
+/clusters/
+fs/
+vpcs/
+runs/
+myvenv/
+__pycache__
+website/ghpcfe/migrations/
+.*.swp
+*.sqlite3
+bin/
+lib/
+share/
+run/
+configuration.yaml
+website/static/
+workbenches/
+dependencies/
+tf/deployment.tar.gz
+key1
+key2
+.terraform/
+.terraform.lock.hcl
+terraform.tfstate*
+terraform.tfvars
+tf/.tkfe.lock
+tf/tfapply.log
+tf/tfdestroy.log

community/front-end/ofe/README.md

@@ -0,0 +1,33 @@
+# Google HPC Toolkit Open Front End
+
+This is a web front-end for HPC applications on GCP. It delegates to the Cloud
+HPC Toolkit to create cloud resources for HPC clusters. Through the convenience
+of a web interface, system administrators can manage the life cycles of HPC
+clusters and install applications; users can prepare & submit HPC jobs and run
+benchmarks. This web application is built upon the Django framework.
+
+## Deployment
+
+This system can be deployed on GCP by an administrator using the following
+steps:
+
+* Arrange a hosting GCP project for this web application.
+* Prepare the client side environment and secure sufficient IAM permissions for
+  the system deployment.
+* When ready, clone this repository and run the deployment script at
+  `hpc-toolkit/community/front-end/deploy.sh` from a client machine or a Cloud
+  Shell. Follow instructions to complete the deployment. The whole process is
+  automated via Terraform and should complete within 15 minutes.
+* Perform post-deployment configurations.
+
+Please visit the [Administrator's Guide](docs/admin_guide.md) for more
+information on system deployment.
+
+Once the deployment is done, the administrator can use the web interface to
+create HPC clusters, install applications, and set up other users. More
+information is available in the [Administrator's Guide](docs/admin_guide.md)
+and [User Guide](docs/user_guide.md).
+
+You are welcome to contribute to this project. The
+[Developer's Guide](docs/developer_guide.md) contains more information on the
+implementation details of the system.