diff --git a/community/examples/hpc-slurm-gromacs.yaml b/community/examples/hpc-slurm-gromacs.yaml index 9a45afc97d..a8d7bb4b27 100644 --- a/community/examples/hpc-slurm-gromacs.yaml +++ b/community/examples/hpc-slurm-gromacs.yaml @@ -29,7 +29,7 @@ deployment_groups: - group: primary modules: - id: network1 - source: modules/network/pre-existing-vpc + source: modules/network/vpc ## Filesystems - id: appsfs @@ -49,6 +49,7 @@ deployment_groups: source: community/modules/scripts/spack-setup settings: install_dir: /sw/spack + spack_ref: v0.20.0 - id: spack-execute source: community/modules/scripts/spack-execute @@ -68,11 +69,11 @@ deployment_groups: projections: all: '{name}/{version}-{compiler.name}-{compiler.version}' commands: | - # Un-comment and update mirror_url to install from spack cache - # if ! spack mirror list | grep -q gcs_cache; then - # spack mirror add --scope site gcs_cache gs://optionally_set_spack_cache_bucket - # fi - # spack buildcache keys --install --trust + ## Un-comment and update mirror_url to install from spack cache + ## if ! spack mirror list | grep -q gcs_cache; then + ## spack mirror add --scope site gcs_cache gs://optionally_set_spack_cache_bucket + ## fi + ## spack buildcache keys --install --trust spack config --scope defaults add config:build_stage:/sw/spack/spack-stage spack config --scope defaults add -f /tmp/projections-config.yaml @@ -82,38 +83,51 @@ deployment_groups: spack compiler find --scope site spack install intel-mpi@2018.4.274%gcc@10.3.0 - spack install gromacs@2023.1 %gcc@10.3.0 ^intel-mpi@2018.4.274 ^cmake@3.26.3 %gcc@4.8.5 + spack install gromacs@2023.1 %gcc@10.3.0 ^intel-mpi@2018.4.274 ^cmake@3.26.3 %gcc@8.5.0 - - id: compute_partition - source: community/modules/compute/SchedMD-slurm-on-gcp-partition - use: - - network1 - - homefs - - appsfs + - id: script + source: modules/scripts/startup-script settings: - partition_name: compute - max_node_count: 20 + runners: + # remove lustre client temporary to avoid startup failure due to known + # issue. + - type: shell + destination: remove_lustre_client.sh + content: | + #!/bin/bash + rm /etc/yum.repos.d/lustre-client.repo + - $(spack-execute.spack_runner) + + - id: compute_nodeset + source: community/modules/compute/schedmd-slurm-gcp-v6-nodeset + use: [network1] + settings: + node_count_dynamic_max: 20 bandwidth_tier: gvnic_enabled - - id: slurm_controller - source: community/modules/scheduler/SchedMD-slurm-on-gcp-controller - use: - - network1 - - homefs - - appsfs - - compute_partition + - id: compute_partition + source: community/modules/compute/schedmd-slurm-gcp-v6-partition + use: [compute_nodeset, homefs, appsfs] settings: - login_node_count: 1 + partition_name: compute - id: slurm_login - source: community/modules/scheduler/SchedMD-slurm-on-gcp-login-node + source: community/modules/scheduler/schedmd-slurm-gcp-v6-login + use: [network1] + settings: + name_prefix: login + machine_type: n2-standard-4 + disable_login_public_ips: false + + - id: slurm_controller + source: community/modules/scheduler/schedmd-slurm-gcp-v6-controller use: - network1 + - compute_partition + - slurm_login - homefs - appsfs - - slurm_controller - - spack-execute settings: - login_machine_type: c2-standard-4 - login_scopes: - - https://www.googleapis.com/auth/cloud-platform + disable_controller_public_ips: false + login_startup_script: $(script.startup_script) + login_startup_scripts_timeout: 21600 diff --git a/tools/cloud-build/daily-tests/ansible_playbooks/test-validation/test-spack.yml b/tools/cloud-build/daily-tests/ansible_playbooks/test-validation/test-spack.yml index b09d1a49a1..67298fe3aa 100644 --- a/tools/cloud-build/daily-tests/ansible_playbooks/test-validation/test-spack.yml +++ b/tools/cloud-build/daily-tests/ansible_playbooks/test-validation/test-spack.yml @@ -18,7 +18,7 @@ ansible.builtin.include_tasks: "tasks/wait-for-startup-script.yml" vars: vm_name: "{{ login_node }}" - timeout_seconds: 7200 + timeout_seconds: 21600 - name: Ensure spack is installed ansible.builtin.command: spack --version changed_when: False diff --git a/tools/cloud-build/daily-tests/tests/spack-gromacs.yml b/tools/cloud-build/daily-tests/tests/spack-gromacs.yml index 412ade50d4..cf548d8c6d 100644 --- a/tools/cloud-build/daily-tests/tests/spack-gromacs.yml +++ b/tools/cloud-build/daily-tests/tests/spack-gromacs.yml @@ -15,14 +15,15 @@ --- test_name: hpc-slurm-gromacs -deployment_name: "spack-gromacs-{{ build }}" +deployment_name: "groma-{{ build }}" +slurm_cluster_name: "groma{{ build[0:5] }}" zone: us-central1-c workspace: /workspace blueprint_yaml: "{{ workspace }}/community/examples/hpc-slurm-gromacs.yaml" -network: "default" +network: "{{ deployment_name }}-net" max_nodes: 5 -login_node: slurm-{{ deployment_name }}-login0 -controller_node: slurm-{{ deployment_name }}-controller +login_node: "{{ slurm_cluster_name }}-login-*" +controller_node: "{{ slurm_cluster_name }}-controller" post_deploy_tests: - test-validation/test-spack.yml custom_vars: