NVIDIA · michael-balint · Jul 2, 2026 · Jul 2, 2026 · Jul 2, 2026
diff --git a/config.example/env.sh b/config.example/env.sh
@@ -4,3 +4,7 @@
 # See deepops/scripts/common.sh for implementation details
 
 DEEPOPS_EXAMPLE_VAR=""
+
+# DeepOps version reported in script debug output.
+# Defaults to the checkout's git tag description; uncomment to override.
+# DEEPOPS_VERSION=""
diff --git a/roles/alertmanager/defaults/main.yml b/roles/alertmanager/defaults/main.yml
@@ -1,6 +1,6 @@
 alertmanager_config_dir: /etc/alertmanager
 alertmanager_config_src: templates/alertmanager.yml.j2
-alertmanager_container: "prom/alertmanager:v0.32.1"
+alertmanager_container: "prom/alertmanager:v0.33.0"
 alertmanager_svc_name: "docker.alertmanager.service"
 alertmanager_docker_volume_name: "deepops_alertmanager_metrics"
 alertmanager_state: started

diff --git a/roles/grafana/defaults/main.yml b/roles/grafana/defaults/main.yml
@@ -2,7 +2,7 @@ grafana_config_dir: /etc/grafana
 grafana_config_template: templates/grafana.ini.j2
 grafana_data_dir: /var/lib/grafana
 grafana_user_id: 472
-grafana_container: "grafana/grafana:13.0.1"
+grafana_container: "grafana/grafana:13.1.0"
 grafana_svc_name: "docker.grafana.service"
 grafana_state: started
 grafana_enabled: yes

diff --git a/roles/nvidia-gpu-operator/defaults/main.yml b/roles/nvidia-gpu-operator/defaults/main.yml
@@ -12,7 +12,7 @@ gpu_operator_nvaie_helm_repo: "https://helm.ngc.nvidia.com/nvaie"
 gpu_operator_nvaie_chart_name: "nvaie/gpu-operator"
 
 # NVAIE GPU Operator may require different version, check NGC enterprise collection.
-gpu_operator_chart_version: "v26.3.1"
+gpu_operator_chart_version: "v26.3.3"
 
 k8s_gpu_mig_strategy: "mixed"
 

diff --git a/roles/nvidia-k8s-gpu-device-plugin/defaults/main.yml b/roles/nvidia-k8s-gpu-device-plugin/defaults/main.yml
@@ -2,6 +2,6 @@
 k8s_gpu_plugin_helm_repo: "https://nvidia.github.io/k8s-device-plugin"
 k8s_gpu_plugin_chart_name: "nvdp/nvidia-device-plugin"
 k8s_gpu_plugin_release_name: "nvidia-device-plugin"
-k8s_gpu_plugin_chart_version: "0.19.1"
+k8s_gpu_plugin_chart_version: "0.19.3"
 k8s_gpu_plugin_init_error: "false"
 k8s_gpu_mig_strategy: "mixed"
diff --git a/roles/nvidia-k8s-gpu-feature-discovery/defaults/main.yml b/roles/nvidia-k8s-gpu-feature-discovery/defaults/main.yml
@@ -2,5 +2,5 @@
 k8s_gpu_feature_discovery_helm_repo: "https://nvidia.github.io/k8s-device-plugin"
 k8s_gpu_feature_discovery_chart_name: "nvgfd/gpu-feature-discovery"
 k8s_gpu_feature_discovery_release_name: "gpu-feature-discovery"
-k8s_gpu_feature_discovery_chart_version: "0.19.1"
+k8s_gpu_feature_discovery_chart_version: "0.19.3"
 k8s_gpu_mig_strategy: "mixed"
diff --git a/roles/nvidia-mig-manager/defaults/main.yml b/roles/nvidia-mig-manager/defaults/main.yml
@@ -1,3 +1,3 @@
 ---
-mig_manager_url_deb: https://github.com/NVIDIA/mig-parted/releases/download/v0.14.1/nvidia-mig-manager_0.14.1-1_amd64.deb
-mig_manager_url_rpm: https://github.com/NVIDIA/mig-parted/releases/download/v0.14.1/nvidia-mig-manager-0.14.1-1.x86_64.rpm
+mig_manager_url_deb: https://github.com/NVIDIA/mig-parted/releases/download/v0.14.2/nvidia-mig-manager_0.14.2-1_amd64.deb
+mig_manager_url_rpm: https://github.com/NVIDIA/mig-parted/releases/download/v0.14.2/nvidia-mig-manager-0.14.2-1.x86_64.rpm
diff --git a/roles/nvidia-network-operator/vars/main.yaml b/roles/nvidia-network-operator/vars/main.yaml
@@ -6,7 +6,7 @@
 # if_name must match k8s network annotation name
 #
 
-nvidia_network_operator_version: "26.1.1"
+nvidia_network_operator_version: "26.4.0"
 nvidia_network_operator_image_tag: "network-operator-v{{ nvidia_network_operator_version }}"
 nvidia_network_operator_namespace: "network-operator"
 nvidia_network_operator_ipam_type: "nv-ipam"

diff --git a/roles/prometheus/defaults/main.yml b/roles/prometheus/defaults/main.yml
@@ -1,7 +1,7 @@
 prometheus_config_dir: /etc/prometheus
 prometheus_config_src: templates/prometheus.yml.j2
 prometheus_alert_rules_src: templates/alert_rules.yml.j2
-prometheus_container: "prom/prometheus:v3.11.3"
+prometheus_container: "prom/prometheus:v3.13.0"
 prometheus_svc_name: "docker.prometheus.service"
 prometheus_docker_volume_name: "deepops_prometheus_metrics"
 prometheus_state: started

diff --git a/roles/slurm/defaults/main.yml b/roles/slurm/defaults/main.yml
@@ -7,7 +7,7 @@ hwloc_build_dir: /opt/deepops/build/hwloc
 pmix_build_dir: /opt/deepops/build/pmix
 
 slurm_workflow_build: yes
-slurm_version: "25.11.6"
+slurm_version: "26.05.1"
 slurm_src_url: "https://download.schedmd.com/slurm/slurm-{{ slurm_version }}.tar.bz2"
 slurm_build_make_clean: no
 slurm_build_dir_cleanup: no

diff --git a/roles/spack/defaults/main.yml b/roles/spack/defaults/main.yml
@@ -1,7 +1,7 @@
 ---
 spack_repo: "https://github.com/spack/spack.git"
 spack_install_dir: "/sw/spack"
-spack_version: "v1.1.1"
+spack_version: "v1.2.0"
 spack_user: "root"
 spack_group: "root"
 

diff --git a/scripts/common.sh b/scripts/common.sh
@@ -19,5 +19,11 @@ else
   source ${DEEPOPS_CONFIG_DIR}/env.sh
 fi
 
+# Determine the DeepOps version for debug output.
+# Can be overridden in config/env.sh; defaults to the checkout's git tag description.
+if [ -z "${DEEPOPS_VERSION}" ]; then
+  DEEPOPS_VERSION="$(git -C "${ROOT_DIR}" describe --tags --always 2>/dev/null || echo unknown)"
+fi
+
 # Print out base debug
 echo "Starting '${0}';  DeepOps version '${DEEPOPS_VERSION}'"
diff --git a/scripts/k8s/deploy_monitoring.sh b/scripts/k8s/deploy_monitoring.sh
@@ -24,7 +24,7 @@ if [ ! -d "${DEEPOPS_CONFIG_DIR}" ]; then
 fi
 
 HELM_CHARTS_REPO_PROMETHEUS="${HELM_CHARTS_REPO_PROMETHEUS:-https://prometheus-community.github.io/helm-charts}"
-HELM_PROMETHEUS_CHART_VERSION="${HELM_PROMETHEUS_CHART_VERSION:-85.0.3}"
+HELM_PROMETHEUS_CHART_VERSION="${HELM_PROMETHEUS_CHART_VERSION:-87.5.1}"
 ingress_name="ingress-nginx"
 
 PROMETHEUS_YAML_CONFIG="${PROMETHEUS_YAML_CONFIG:-${DEEPOPS_CONFIG_DIR}/helm/monitoring.yml}"