Upgrade dependencies

hanwen-cluster · gmarciani · commit 6ef0b0b974e3 · 2025-08-08T08:02:32.000-04:00
- Upgrade Slurm to version 24.11.6 (from 24.05.8). - Upgrade EFA installer to 1.42.0 (from 1.41.0). - Efa-driver: efa-2.15.3-1 - Efa-config: efa-config-1.18-1 - Efa-profile: efa-profile-1.7-1 - Libfabric-aws: libfabric-aws-2.1.0-3 - Rdma-core: rdma-core-57.0-1 - Open MPI: openmpi40-aws-4.1.7-2 and openmpi50-aws-5.0.6-11 - Upgrade Cinc Client to version to 18.4.12 from 18.2.7. - Upgrade NVIDIA driver to version 570.172.08 (from 570.86.15) for all OSs except AL2. - Upgrade CUDA Toolkit to version 12.8.1 (from 12.8.0) for all OSs except AL2. - Upgrade DCGM to version 4.2.3 (from 3.3.6) for all OSs except AL2. - Upgrade Python to 3.12.11 (from 3.12.8) for all OSs except AL2. - Upgrade Intel MPI Library to 2021.16.0 (from 2021.13.1). Among the above upgrade, DCGM is a major version upgrade (from version 3 to version 4) This is a new change in DCGM 4: ``` Installation assets are no longer shipped in a single monolithic package. Instead, installation assets have been split among several packages, allowing clients to opt-out of the installation of assets not applicable to their use case. Component packages are as follows: datacenter-gpu-manager-4-core Provides nv-hostengine binary and other CUDA-agnostic installation assets available through the DCGM open source product datacenter-gpu-manager-4-cuda11 Provides the CUDA11-specific binaries available through the DCGM open source product datacenter-gpu-manager-4-cuda12 Provides the CUDA12-specific binaries available through the DCGM open source product datacenter-gpu-manager-4-proprietary Provides CUDA-agnostic installation assets not distributed as part of the DCGM open source product datacenter-gpu-manager-4-proprietary-cuda11 Provides CUDA11 binaries not distributed as part of the DCGM open source product datacenter-gpu-manager-4-proprietary-cuda12 Provides CUDA12 binaries not distributed as part of the DCGM open source product datacenter-gpu-manager-4-development Provides files necessary for the development of downstream software dependent on the DCGM library ``` https://docs.nvidia.com/datacenter/dcgm/latest/release-notes/changelog.html Signed-off-by: Hanwen <hanwenli@amazon.com> Signed-off-by: Hanwen <hanwenli@amazon.com>
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -11,7 +11,20 @@ This file is used to list changes made in each version of the AWS ParallelCluste
 
 **CHANGES**
 - Ubuntu 20.04 is no longer supported.
-- Upgrade Slurm to version 24.11.5.
+- Upgrade Slurm to version 24.11.6 (from 24.05.8).
+- Upgrade EFA installer to 1.42.0 (from 1.41.0).
+  - Efa-driver: efa-2.15.3-1
+  - Efa-config: efa-config-1.18-1
+  - Efa-profile: efa-profile-1.7-1
+  - Libfabric-aws: libfabric-aws-2.1.0-3
+  - Rdma-core: rdma-core-57.0-1
+  - Open MPI: openmpi40-aws-4.1.7-2 and openmpi50-aws-5.0.6-11
+- Upgrade Cinc Client to version to 18.4.12 from 18.2.7.
+- Upgrade NVIDIA driver to version 570.172.08 (from 570.86.15) for all OSs except AL2.
+- Upgrade CUDA Toolkit to version 12.8.1 (from 12.8.0) for all OSs except AL2.
+- Upgrade DCGM to version 4.2.3 (from 3.3.6) for all OSs except AL2.
+- Upgrade Python to 3.12.11 (from 3.12.8) for all OSs except AL2.
+- Upgrade Intel MPI Library to 2021.16.0 (from 2021.13.1).
 - Addressed cluster id mismatch known issue by deleting the file `/var/spool/slurm.state/clustername` before configuring Slurm accounting.
 - Upgrade DCV to version 2024.0-19030.
 - Remove `berkshelf`. All cookbooks are local and do not need `berkshelf` dependency management.
diff --git a/cookbooks/aws-parallelcluster-awsbatch/test/controls/awsbatch_virtualenv_spec.rb b/cookbooks/aws-parallelcluster-awsbatch/test/controls/awsbatch_virtualenv_spec.rb
@@ -13,7 +13,7 @@
 pyenv_dir = "#{base_dir}/pyenv"
 
 control 'tag:install_awsbatch_virtualenv_created' do
-  python_version = os_properties.alinux2? ? '3.9.20' : '3.12.8'
+  python_version = os_properties.alinux2? ? '3.9.20' : '3.12.11'
   title "awsbatch virtualenv should be created on #{python_version}"
   only_if { !os_properties.redhat? }
 
diff --git a/cookbooks/aws-parallelcluster-computefleet/kitchen.computefleet-config.yml b/cookbooks/aws-parallelcluster-computefleet/kitchen.computefleet-config.yml
@@ -31,7 +31,7 @@ suites:
     attributes:
       cluster:
         custom_node_package: https://github.com/aws/aws-parallelcluster-node/archive/develop.tar.gz
-        python-version: 3.12.8
+        python-version: 3.12.11
         node_virtualenv_path: /opt/parallelcluster/pyenv/versions/node_virtualenv
   - name: fleet_status
     run_list:
diff --git a/cookbooks/aws-parallelcluster-entrypoints/Berksfile.lock b/cookbooks/aws-parallelcluster-entrypoints/Berksfile.lock
@@ -0,0 +1,84 @@
+DEPENDENCIES
+  aws-parallelcluster-awsbatch
+    path: ../aws-parallelcluster-awsbatch
+  aws-parallelcluster-computefleet
+    path: ../aws-parallelcluster-computefleet
+  aws-parallelcluster-entrypoints
+    path: .
+    metadata: true
+  aws-parallelcluster-environment
+    path: ../aws-parallelcluster-environment
+  aws-parallelcluster-platform
+    path: ../aws-parallelcluster-platform
+  aws-parallelcluster-shared
+    path: ../aws-parallelcluster-shared
+  aws-parallelcluster-slurm
+    path: ../aws-parallelcluster-slurm
+  aws-parallelcluster-tests
+    path: ../aws-parallelcluster-tests
+  iptables
+    path: ../third-party/iptables-8.0.0
+  line
+    path: ../third-party/line-4.5.21
+  nfs
+    path: ../third-party/nfs-5.1.5
+  openssh
+    path: ../third-party/openssh-2.11.14
+  yum
+    path: ../third-party/yum-7.4.20
+  yum-epel
+    path: ../third-party/yum-epel-5.0.8
+
+GRAPH
+  aws-parallelcluster-awsbatch (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    iptables (~> 8.0.0)
+    line (~> 4.5.21)
+    nfs (~> 5.1.5)
+    openssh (~> 2.11.14)
+    yum (~> 7.4.20)
+    yum-epel (~> 5.0.8)
+  aws-parallelcluster-computefleet (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+  aws-parallelcluster-entrypoints (3.13.0)
+    aws-parallelcluster-awsbatch (~> 3.13.0)
+    aws-parallelcluster-computefleet (~> 3.13.0)
+    aws-parallelcluster-environment (~> 3.13.0)
+    aws-parallelcluster-platform (~> 3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    aws-parallelcluster-slurm (~> 3.13.0)
+  aws-parallelcluster-environment (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    line (~> 4.5.21)
+    nfs (~> 5.1.5)
+  aws-parallelcluster-platform (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    line (~> 4.5.21)
+  aws-parallelcluster-shared (3.13.0)
+    yum (~> 7.4.20)
+    yum-epel (~> 5.0.8)
+  aws-parallelcluster-slurm (3.13.0)
+    aws-parallelcluster-computefleet (~> 3.13.0)
+    aws-parallelcluster-environment (~> 3.13.0)
+    aws-parallelcluster-platform (~> 3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    iptables (~> 8.0.0)
+    line (~> 4.5.21)
+    nfs (~> 5.1.5)
+    openssh (~> 2.11.14)
+    yum (~> 7.4.20)
+    yum-epel (~> 5.0.8)
+  aws-parallelcluster-tests (3.13.0)
+    aws-parallelcluster-computefleet (~> 3.13.0)
+    aws-parallelcluster-environment (~> 3.13.0)
+    aws-parallelcluster-platform (~> 3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    aws-parallelcluster-slurm (~> 3.13.0)
+  iptables (8.0.0)
+  line (4.5.21)
+  nfs (5.1.5)
+    line (>= 0.0.0)
+  openssh (2.11.14)
+    iptables (>= 7.0)
+  yum (7.4.20)
+  yum-epel (5.0.8)
diff --git a/cookbooks/aws-parallelcluster-environment/Berksfile.lock b/cookbooks/aws-parallelcluster-environment/Berksfile.lock
@@ -0,0 +1,28 @@
+DEPENDENCIES
+  aws-parallelcluster-environment
+    path: .
+    metadata: true
+  aws-parallelcluster-shared
+    path: ../aws-parallelcluster-shared
+  line
+    path: ../third-party/line-4.5.21
+  nfs
+    path: ../third-party/nfs-5.1.5
+  yum
+    path: ../third-party/yum-7.4.20
+  yum-epel
+    path: ../third-party/yum-epel-5.0.8
+
+GRAPH
+  aws-parallelcluster-environment (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    line (~> 4.5.21)
+    nfs (~> 5.1.5)
+  aws-parallelcluster-shared (3.13.0)
+    yum (~> 7.4.20)
+    yum-epel (~> 5.0.8)
+  line (4.5.21)
+  nfs (5.1.5)
+    line (>= 0.0.0)
+  yum (7.4.20)
+  yum-epel (5.0.8)
diff --git a/cookbooks/aws-parallelcluster-environment/attributes/environment.rb b/cookbooks/aws-parallelcluster-environment/attributes/environment.rb
@@ -70,8 +70,8 @@
 
 default['cluster']['head_node_private_ip'] = nil
 
-default['cluster']['efa']['version'] = '1.41.0'
-default['cluster']['efa']['sha256'] = '3506354cdfbe31ff552fe75f5d0d9bb7efd29cf79bd99457347d29c751c38f9f'
+default['cluster']['efa']['version'] = '1.42.0'
+default['cluster']['efa']['sha256'] = '4114fe612905ee05083ae5cb391a00a012510f3abfecc642d86c9a5ae4be9008'
 
 default['cluster']['efs']['version'] = '2.3.1'
 default['cluster']['efs']['sha256'] = 'ced12f82e76f9740476b63f30c49bd76cc00b6375e12a9f5f7ba852635c49e15'
diff --git a/cookbooks/aws-parallelcluster-environment/spec/unit/resources/efa_spec.rb b/cookbooks/aws-parallelcluster-environment/spec/unit/resources/efa_spec.rb
@@ -2,8 +2,8 @@
 
 # parallelcluster default source dir defined in attributes
 source_dir = '/opt/parallelcluster/sources'
-efa_version = '1.41.0'
-efa_checksum = '3506354cdfbe31ff552fe75f5d0d9bb7efd29cf79bd99457347d29c751c38f9f'
+efa_version = '1.42.0'
+efa_checksum = '4114fe612905ee05083ae5cb391a00a012510f3abfecc642d86c9a5ae4be9008'
 
 class ConvergeEfa
   def self.setup(chef_run, efa_version: nil, efa_checksum: nil)
diff --git a/cookbooks/aws-parallelcluster-environment/test/controls/cfn_bootstrap_spec.rb b/cookbooks/aws-parallelcluster-environment/test/controls/cfn_bootstrap_spec.rb
@@ -13,7 +13,7 @@
 pyenv_dir = "#{base_dir}/pyenv"
 
 control 'tag:install_cfnbootstrap_virtualenv_created' do
-  cfn_python_version = os_properties.alinux2? ? '3.9.20' : '3.12.8'
+  cfn_python_version = os_properties.alinux2? ? '3.9.20' : '3.12.11'
   title "cfnbootstrap virtualenv should be created on #{cfn_python_version}"
   only_if { !os_properties.redhat_on_docker? }
 
diff --git a/cookbooks/aws-parallelcluster-platform/Berksfile.lock b/cookbooks/aws-parallelcluster-platform/Berksfile.lock
@@ -0,0 +1,23 @@
+DEPENDENCIES
+  aws-parallelcluster-platform
+    path: .
+    metadata: true
+  aws-parallelcluster-shared
+    path: ../aws-parallelcluster-shared
+  line
+    path: ../third-party/line-4.5.21
+  yum
+    path: ../third-party/yum-7.4.20
+  yum-epel
+    path: ../third-party/yum-epel-5.0.8
+
+GRAPH
+  aws-parallelcluster-platform (3.13.0)
+    aws-parallelcluster-shared (~> 3.13.0)
+    line (~> 4.5.21)
+  aws-parallelcluster-shared (3.13.0)
+    yum (~> 7.4.20)
+    yum-epel (~> 5.0.8)
+  line (4.5.21)
+  yum (7.4.20)
+  yum-epel (5.0.8)
diff --git a/cookbooks/aws-parallelcluster-platform/attributes/platform.rb b/cookbooks/aws-parallelcluster-platform/attributes/platform.rb
@@ -17,9 +17,10 @@
 # NVidia
 default['cluster']['nvidia']['enabled'] = 'no'
 default['cluster']['nvidia']['driver_version'] = '570.172.08'
-default['cluster']['nvidia']['dcgm_version'] = '3.3.6'
+default['cluster']['nvidia']['dcgm_version'] = '4.2.3-2'
 if platform?('amazon') && node['platform_version'] == "2"
   default['cluster']['nvidia']['driver_version'] = '550.127.08'
+  default['cluster']['nvidia']['dcgm_version'] = '3.3.6-1'
 end
 
 # nvidia-imex
diff --git a/cookbooks/aws-parallelcluster-platform/recipes/install/cuda.rb b/cookbooks/aws-parallelcluster-platform/recipes/install/cuda.rb
@@ -20,9 +20,9 @@
 # Cuda installer from https://developer.nvidia.com/cuda-toolkit-archive
 # Cuda installer naming: cuda_11.8.0_520.61.05_linux
 cuda_version = '12.8'
-cuda_patch = '0'
+cuda_patch = '1'
 cuda_complete_version = "#{cuda_version}.#{cuda_patch}"
-cuda_version_suffix = '570.86.10'
+cuda_version_suffix = '570.124.06'
 cuda_samples_version = '12.8'
 if platform?('amazon') && node['platform_version'] == "2"
   cuda_version = '12.4'
diff --git a/cookbooks/aws-parallelcluster-platform/recipes/install/intel_mpi.rb b/cookbooks/aws-parallelcluster-platform/recipes/install/intel_mpi.rb
@@ -16,7 +16,7 @@
 # limitations under the License.
 
 intelmpi_supported = !arm_instance?
-intelmpi_version = '2021.13'
+intelmpi_version = '2021.16'
 
 node.default['conditions']['intel_mpi_supported'] = intelmpi_supported
 node.default['cluster']['intelmpi']['version'] = intelmpi_version
@@ -25,9 +25,9 @@
 
 return unless intelmpi_supported
 
-intelmpi_full_version = "#{intelmpi_version}.1.769"
+intelmpi_full_version = "#{intelmpi_version}.0.443"
 intelmpi_installation_path = "/opt/intel/mpi/#{intelmpi_version}"
-intelmpi_installer = "l_mpi_oneapi_p_#{intelmpi_full_version}_offline.sh"
+intelmpi_installer = "intel-mpi-#{intelmpi_full_version}_offline.sh"
 intelmpi_installer_path = "#{node['cluster']['sources_dir']}/#{intelmpi_installer}"
 intelmpi_installer_url = "#{node['cluster']['artifacts_s3_url']}/impi/#{intelmpi_installer}"
 intelmpi_qt_version = '6.5.3'
diff --git a/cookbooks/aws-parallelcluster-platform/resources/nvidia_dcgm/partial/_nvidia_dcgm_debian.rb b/cookbooks/aws-parallelcluster-platform/resources/nvidia_dcgm/partial/_nvidia_dcgm_debian.rb
@@ -13,34 +13,45 @@
 # See the License for the specific language governing permissions and limitations under the License.
 
 action :install_package do
-  remote_file "#{node['cluster']['sources_dir']}/#{dcgm_package}-#{package_version}.deb" do
-    source "#{dcgm_url}"
-    mode '0644'
-    retries 3
-    retry_delay 5
-    action :create_if_missing
-  end
+  packages_urls_list = if package_version.start_with?("3.")
+                         [dcgm_package]
+                       else
+                         [dcgm4_core_package, dcgm4_package]
+                       end
+  packages_urls_list.each do |package|
+    remote_file "#{node['cluster']['sources_dir']}/#{package}-#{package_version}.deb" do
+      source "#{node['cluster']['artifacts_s3_url']}/dependencies/nvidia_dcgm/#{platform}/#{package}_#{package_version}_#{arch_suffix}.deb"
+      mode '0644'
+      retries 3
+      retry_delay 5
+      action :create_if_missing
+    end
 
-  bash "Install #{dcgm_package}" do
-    user 'root'
-    cwd node['cluster']['sources_dir']
-    code <<-DCGM_INSTALL
-    set -e
-    dpkg -i #{dcgm_package}-#{package_version}.deb
-    DCGM_INSTALL
-    retries 3
-    retry_delay 5
+    bash "Install #{package}" do
+      user 'root'
+      cwd node['cluster']['sources_dir']
+      code <<-DCGM_INSTALL
+      set -e
+      dpkg -i #{package}-#{package_version}.deb
+      DCGM_INSTALL
+      retries 3
+      retry_delay 5
+    end
   end
 end
 
-def dcgm_url
-  "#{node['cluster']['artifacts_s3_url']}/dependencies/nvidia_dcgm/#{platform}/#{dcgm_package}_#{package_version}_#{arch_suffix}.deb"
-end
-
 def dcgm_package
   'datacenter-gpu-manager'
 end
 
+def dcgm4_package
+  "#{dcgm_package}-4-cuda12"
+end
+
+def dcgm4_core_package
+  "#{dcgm_package}-4-core"
+end
+
 def arch_suffix
   arm_instance? ? 'arm64' : 'amd64'
 end
diff --git a/cookbooks/aws-parallelcluster-platform/resources/nvidia_dcgm/partial/_nvidia_dcgm_rhel.rb b/cookbooks/aws-parallelcluster-platform/resources/nvidia_dcgm/partial/_nvidia_dcgm_rhel.rb
@@ -13,34 +13,47 @@
 # See the License for the specific language governing permissions and limitations under the License.
 
 action :install_package do
-  remote_file "#{node['cluster']['sources_dir']}/#{dcgm_package}-#{package_version}.rpm" do
-    source "#{dcgm_url}"
-    mode '0644'
-    retries 3
-    retry_delay 5
-    action :create_if_missing
+  if package_version.start_with?("3.")
+    packages_urls_list = [dcgm_package]
+    package_url_separator = "-"
+  else
+    packages_urls_list = [dcgm4_core_package, dcgm4_package]
+    package_url_separator = "."
   end
+  packages_urls_list.each do |package|
+    remote_file "#{node['cluster']['sources_dir']}/#{package}-#{package_version}.rpm" do
+      source "#{node['cluster']['artifacts_s3_url']}/dependencies/nvidia_dcgm/#{platform}/#{package}-#{package_version}#{package_url_separator}#{arch_suffix}.rpm"
+      mode '0644'
+      retries 3
+      retry_delay 5
+      action :create_if_missing
+    end
 
-  bash "Install #{dcgm_package}" do
-    user 'root'
-    cwd node['cluster']['sources_dir']
-    code <<-DCGM_INSTALL
-    set -e
-    yum install -y #{dcgm_package}-#{package_version}.rpm
-    DCGM_INSTALL
-    retries 3
-    retry_delay 5
+    bash "Install #{package}" do
+      user 'root'
+      cwd node['cluster']['sources_dir']
+      code <<-DCGM_INSTALL
+      set -e
+      yum install -y #{package}-#{package_version}.rpm
+      DCGM_INSTALL
+      retries 3
+      retry_delay 5
+    end
   end
 end
 
-def dcgm_url
-  "#{node['cluster']['artifacts_s3_url']}/dependencies/nvidia_dcgm/#{platform}/#{dcgm_package}-#{package_version}-1-#{arch_suffix}.rpm"
-end
-
 def dcgm_package
   'datacenter-gpu-manager'
 end
 
+def dcgm4_package
+  "#{dcgm_package}-4-cuda12"
+end
+
+def dcgm4_core_package
+  "#{dcgm_package}-4-core"
+end
+
 def arch_suffix
   arm_instance? ? 'aarch64' : 'x86_64'
 end
diff --git a/cookbooks/aws-parallelcluster-platform/spec/unit/recipes/cuda_spec.rb b/cookbooks/aws-parallelcluster-platform/spec/unit/recipes/cuda_spec.rb
@@ -2,9 +2,9 @@
 
 describe 'aws-parallelcluster-platform::cuda' do
   cached(:cuda_version) { '12.8' }
-  cached(:cuda_patch) { '0' }
+  cached(:cuda_patch) { '1' }
   cached(:cuda_complete_version) { "#{cuda_version}.#{cuda_patch}" }
-  cached(:cuda_version_suffix) { '570.86.10' }
+  cached(:cuda_version_suffix) { '570.124.06' }
 
   context 'when nvidia not enabled' do
     cached(:chef_run) do
diff --git a/cookbooks/aws-parallelcluster-platform/spec/unit/recipes/intel_mpi_spec.rb b/cookbooks/aws-parallelcluster-platform/spec/unit/recipes/intel_mpi_spec.rb
diff --git a/cookbooks/aws-parallelcluster-platform/spec/unit/resources/nvidia_dcgm_spec.rb b/cookbooks/aws-parallelcluster-platform/spec/unit/resources/nvidia_dcgm_spec.rb
diff --git a/cookbooks/aws-parallelcluster-platform/test/controls/nvidia_dcgm_spec.rb b/cookbooks/aws-parallelcluster-platform/test/controls/nvidia_dcgm_spec.rb
diff --git a/cookbooks/aws-parallelcluster-shared/Berksfile.lock b/cookbooks/aws-parallelcluster-shared/Berksfile.lock
diff --git a/cookbooks/aws-parallelcluster-shared/attributes/versions.rb b/cookbooks/aws-parallelcluster-shared/attributes/versions.rb
diff --git a/cookbooks/aws-parallelcluster-slurm/Berksfile.lock b/cookbooks/aws-parallelcluster-slurm/Berksfile.lock
diff --git a/cookbooks/aws-parallelcluster-slurm/attributes/versions.rb b/cookbooks/aws-parallelcluster-slurm/attributes/versions.rb
