├── .gitignore ├── _static └── custom.css ├── docs ├── index.html ├── images │ ├── Ecosystem.png │ ├── gaudinet_image.png │ ├── model-server-status.png │ ├── rhoai-deploy-model.png │ ├── tgi-serving-runtime.png │ ├── rhods-dashboard-route.png │ ├── Ovms-Gpu-resource-limit.png │ ├── Ovms-Gpu-resource-request.png │ ├── openvino-accelerator-field.png │ ├── Operator-Architecture-Options.png │ ├── accelerator-profile-dropdown.png │ ├── congestion_test_single_leaf_switch.png │ ├── congestion_test_single_leaf_spine_switch.png │ ├── bisection_bandwidth_testing_on_leaf_switch.png │ ├── bisection_bandwidth_testing_on_leaf_spine_switchs.png │ ├── Intel-Technology-Enabling-for-OpenShift-Architecture.png │ ├── bisection_bandwidth_testing_all_Gaudis_leaf_spine_switchs.png │ ├── bisection_bandwidth_testing_all_nodes_leaf_spine_switchs.png │ └── Intel-Enterprise-AI-Foundation-for-OpenShift-Training-Solution.png ├── releases.rst └── supported_platforms.md ├── requirements.txt ├── security ├── dsa_serviceAccount.yaml ├── dsa_role.yaml ├── dsa_roleBinding.yaml ├── qatlib_rbac.yaml ├── dsa_scc.yaml └── qatlib_scc.yaml ├── tests ├── l2 │ ├── dsa │ │ ├── dsa_imagestream.yaml │ │ ├── rh_auth.yaml │ │ ├── dsa_job.yaml │ │ ├── dsa_build.yaml │ │ └── README.md │ ├── sgx │ │ ├── sgx_imagestream.yaml │ │ ├── sgx_job.yaml │ │ ├── README.md │ │ └── sgx_build.yaml │ ├── dgpu │ │ ├── clinfo_job.yaml │ │ ├── intelvpl_job.yaml │ │ ├── hwinfo_job.yaml │ │ ├── vainfo_job.yaml │ │ ├── hwinfo_build.yaml │ │ ├── clinfo_build.yaml │ │ ├── vainfo_build.yaml │ │ ├── intelvpl_build.yaml │ │ └── README.md │ ├── qat │ │ ├── qatlib_job.yaml │ │ ├── qatlib_build.yaml │ │ └── README.md │ └── README.md ├── gaudi │ └── l2 │ │ ├── vllm_hf_secret.yaml │ │ ├── test-pod.yaml │ │ ├── hl-smi_job.yaml │ │ ├── hl-smi-firmware_job.yaml │ │ ├── hccl_build.yaml │ │ ├── hccl_job.yaml │ │ ├── vllm_deployment.yaml │ │ ├── vllm_buildconfig.yaml │ │ └── README.md ├── l3 │ └── README.md └── README.md ├── e2e └── inference │ ├── accelerator_profile_gaudi.yaml │ ├── accelerator_profile_flex140.yaml │ └── README.md ├── security.md ├── nfd ├── node-feature-discovery-openshift.yaml ├── node-feature-rules-openshift.yaml └── README.md ├── device_plugins ├── sgx_device_plugin.yaml ├── gpu_device_plugin.yaml ├── dsa_device_plugin.yaml ├── qat_device_plugin.yaml ├── install_operator.yaml ├── deploy_sgx.md ├── deploy_gpu.md ├── deploy_dsa.md ├── README.md └── deploy_qat.md ├── machine_configuration ├── 100-intel-iommu-on.yaml └── README.md ├── gaudi ├── gaudi_spi_fw_upgrade_job.yaml ├── gaudi_install_operator.yaml ├── gaudi_spi_fw_upgrade_build.yaml ├── Gaudi-SPI-Firmware-Upgrade.md ├── gaudi_cluster_policy.yaml └── README.md ├── workloads └── opea │ └── chatqna │ ├── create_megaservice_container.sh │ ├── persistent_volumes.yaml │ ├── tgi_gaudi_servingruntime.yaml │ ├── chatqna_megaservice_buildconfig.yaml │ ├── redis_deployment_service.yaml │ ├── chatqna_megaservice_deployment.yaml │ └── README.md ├── _templates └── versions.html ├── set-version.sh ├── kmmo ├── intel-dgpu.yaml ├── intel-dgpu-on-premise-build.yaml └── README.md ├── make.bat ├── playbooks ├── install_device_plugins_operator.yaml ├── README.md ├── install_nfd_operator.yaml ├── configure_nfd.yaml └── intel_ocp_provisioning.yaml ├── index.rst ├── Makefile ├── CONTRIBUTING.md ├── one_click ├── README.md ├── gaudi_provisioning_playbook.yaml └── gpu_provisioning_playbook.yaml ├── .github └── workflows │ └── publish.yml ├── conf.py └── CODE_OF_CONDUCT.md /.gitignore: -------------------------------------------------------------------------------- 1 | _build 2 | -------------------------------------------------------------------------------- /_static/custom.css: -------------------------------------------------------------------------------- 1 | .wy-nav-content { 2 | max-width: 90% !important; 3 | } -------------------------------------------------------------------------------- /docs/index.html: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /docs/images/Ecosystem.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Ecosystem.png -------------------------------------------------------------------------------- /docs/images/gaudinet_image.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/gaudinet_image.png -------------------------------------------------------------------------------- /docs/images/model-server-status.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/model-server-status.png -------------------------------------------------------------------------------- /docs/images/rhoai-deploy-model.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/rhoai-deploy-model.png -------------------------------------------------------------------------------- /docs/images/tgi-serving-runtime.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/tgi-serving-runtime.png -------------------------------------------------------------------------------- /docs/images/rhods-dashboard-route.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/rhods-dashboard-route.png -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | docutils<0.18 2 | sphinx 3 | sphinx_rtd_theme 4 | recommonmark 5 | sphinx-markdown-tables 6 | sphinx-md 7 | myst_parser 8 | GitPython -------------------------------------------------------------------------------- /docs/images/Ovms-Gpu-resource-limit.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Ovms-Gpu-resource-limit.png -------------------------------------------------------------------------------- /docs/images/Ovms-Gpu-resource-request.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Ovms-Gpu-resource-request.png -------------------------------------------------------------------------------- /docs/images/openvino-accelerator-field.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/openvino-accelerator-field.png -------------------------------------------------------------------------------- /docs/images/Operator-Architecture-Options.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Operator-Architecture-Options.png -------------------------------------------------------------------------------- /docs/images/accelerator-profile-dropdown.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/accelerator-profile-dropdown.png -------------------------------------------------------------------------------- /docs/images/congestion_test_single_leaf_switch.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/congestion_test_single_leaf_switch.png -------------------------------------------------------------------------------- /docs/images/congestion_test_single_leaf_spine_switch.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/congestion_test_single_leaf_spine_switch.png -------------------------------------------------------------------------------- /docs/images/bisection_bandwidth_testing_on_leaf_switch.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/bisection_bandwidth_testing_on_leaf_switch.png -------------------------------------------------------------------------------- /docs/images/bisection_bandwidth_testing_on_leaf_spine_switchs.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/bisection_bandwidth_testing_on_leaf_spine_switchs.png -------------------------------------------------------------------------------- /docs/images/Intel-Technology-Enabling-for-OpenShift-Architecture.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Intel-Technology-Enabling-for-OpenShift-Architecture.png -------------------------------------------------------------------------------- /docs/images/bisection_bandwidth_testing_all_Gaudis_leaf_spine_switchs.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/bisection_bandwidth_testing_all_Gaudis_leaf_spine_switchs.png -------------------------------------------------------------------------------- /docs/images/bisection_bandwidth_testing_all_nodes_leaf_spine_switchs.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/bisection_bandwidth_testing_all_nodes_leaf_spine_switchs.png -------------------------------------------------------------------------------- /security/dsa_serviceAccount.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: v1 5 | kind: ServiceAccount 6 | metadata: 7 | name: intel-dsa 8 | namespace: intel-dsa -------------------------------------------------------------------------------- /docs/images/Intel-Enterprise-AI-Foundation-for-OpenShift-Training-Solution.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/HEAD/docs/images/Intel-Enterprise-AI-Foundation-for-OpenShift-Training-Solution.png -------------------------------------------------------------------------------- /tests/l2/dsa/dsa_imagestream.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-dsa-workload 8 | namespace: intel-dsa -------------------------------------------------------------------------------- /tests/l2/sgx/sgx_imagestream.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-sgx-workload 8 | namespace: intel-sgx -------------------------------------------------------------------------------- /tests/gaudi/l2/vllm_hf_secret.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | apiVersion: v1 4 | kind: Secret 5 | metadata: 6 | name: hf-token 7 | namespace: gaudi-validation 8 | type: Opaque 9 | data: 10 | hf-token: # Add your token -------------------------------------------------------------------------------- /tests/l2/dsa/rh_auth.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: v1 5 | kind: Secret 6 | metadata: 7 | name: rh-auth 8 | namespace: intel-dsa 9 | type: Opaque 10 | data: 11 | username: # Add username 12 | password: # Add password 13 | -------------------------------------------------------------------------------- /e2e/inference/accelerator_profile_gaudi.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | apiVersion: dashboard.opendatahub.io/v1 4 | kind: AcceleratorProfile 5 | metadata: 6 | name: intel-gaudi-ai-accelerator 7 | spec: 8 | displayName: Intel Gaudi AI Accelerator 9 | description: Intel Gaudi AI Accelerator 10 | enabled: true 11 | identifier: habana.ai/gaudi -------------------------------------------------------------------------------- /security/dsa_role.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: rbac.authorization.k8s.io/v1 5 | kind: Role 6 | metadata: 7 | name: intel-dsa 8 | namespace: intel-dsa 9 | rules: 10 | - apiGroups: 11 | - security.openshift.io 12 | resources: 13 | - securitycontextconstraints 14 | resourceNames: 15 | - intel-dsa-scc 16 | verbs: 17 | - use -------------------------------------------------------------------------------- /security.md: -------------------------------------------------------------------------------- 1 | # Security Policy 2 | Intel is committed to rapidly addressing security vulnerabilities affecting our customers and providing clear guidance on the solution, impact, severity and mitigation. 3 | 4 | ## Reporting a Vulnerability 5 | Please report any security vulnerabilities in this project utilizing the guidelines [here](https://www.intel.com/content/www/us/en/security-center/vulnerability-handling-guidelines.html). 6 | -------------------------------------------------------------------------------- /security/dsa_roleBinding.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: rbac.authorization.k8s.io/v1 5 | kind: RoleBinding 6 | metadata: 7 | name: intel-dsa 8 | namespace: intel-dsa 9 | roleRef: 10 | apiGroup: rbac.authorization.k8s.io 11 | kind: Role 12 | name: intel-dsa 13 | subjects: 14 | - kind: ServiceAccount 15 | name: intel-dsa 16 | namespace: intel-dsa -------------------------------------------------------------------------------- /tests/gaudi/l2/test-pod.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | apiVersion: v1 4 | kind: Pod 5 | metadata: 6 | name: test 7 | labels: 8 | app: test 9 | namespace: gaudi-validation 10 | spec: 11 | containers: 12 | - name: test 13 | command: [ "/bin/bash", "-c", "--" ] 14 | args: [ "while true; do sleep 30; done;"] 15 | image: registry.access.redhat.com/ubi9-minimal:latest 16 | -------------------------------------------------------------------------------- /nfd/node-feature-discovery-openshift.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: nfd.openshift.io/v1 5 | kind: NodeFeatureDiscovery 6 | metadata: 7 | name: nfd-instance 8 | namespace: openshift-nfd 9 | spec: 10 | operand: 11 | image: registry.redhat.io/openshift4/ose-node-feature-discovery-rhel9:v4.18 12 | imagePullPolicy: Always 13 | servicePort: 12000 14 | workerConfig: 15 | configData: | -------------------------------------------------------------------------------- /device_plugins/sgx_device_plugin.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: deviceplugin.intel.com/v1 5 | kind: SgxDevicePlugin 6 | metadata: 7 | name: sgxdeviceplugin-sample 8 | spec: 9 | image: registry.connect.redhat.com/intel/intel-sgx-plugin@sha256:f2c77521c6dae6b4db1896a5784ba8b06a5ebb2a01684184fc90143cfcca7bf4 10 | enclaveLimit: 110 11 | provisionLimit: 110 12 | logLevel: 4 13 | nodeSelector: 14 | intel.feature.node.kubernetes.io/sgx: "true" -------------------------------------------------------------------------------- /device_plugins/gpu_device_plugin.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: deviceplugin.intel.com/v1 5 | kind: GpuDevicePlugin 6 | metadata: 7 | name: gpudeviceplugin-sample 8 | spec: 9 | image: registry.connect.redhat.com/intel/intel-gpu-plugin@sha256:e2c2ce658e78c35c425f16a4f8e85c5f32ce31848d9b53a644a05e7f8b7f71b0 10 | preferredAllocationPolicy: none 11 | sharedDevNum: 1 12 | logLevel: 4 13 | nodeSelector: 14 | intel.feature.node.kubernetes.io/gpu: "true" -------------------------------------------------------------------------------- /e2e/inference/accelerator_profile_flex140.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | apiVersion: dashboard.opendatahub.io/v1 4 | kind: AcceleratorProfile 5 | metadata: 6 | name: intel-gpu-flex-series-140 7 | spec: 8 | displayName: Intel® Data Center GPU Flex Series 140 9 | description: Intel Data Center GPU for inference 10 | enabled: true 11 | identifier: gpu.intel.com/i915 12 | tolerations: 13 | - effect: NoSchedule 14 | key: gpu.intel.com/flex-140 15 | operator: Exists -------------------------------------------------------------------------------- /tests/l3/README.md: -------------------------------------------------------------------------------- 1 | # L3 Test 2 | 3 | ```{admonition} Disclaimer 4 | Please note that this module is currently under development and may contain partially implemented features, therefore it is not supported in the current release. 5 | ``` 6 | 7 | ## L3 Overview 8 | Layer 3 consists of the software stack used to provision the e2e test for specific Intel features like Intel® Data Center GPU Flex Series and Intel® SGX. This layer is directly relied by the reference workloads -layer 4. 9 | 10 | ### GPU 11 | Please follow the [README](/e2e/inference/README.md) for Intel Data Center GPU Flex Series specific prerequisites. -------------------------------------------------------------------------------- /machine_configuration/100-intel-iommu-on.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: machineconfiguration.openshift.io/v1 5 | kind: MachineConfig 6 | metadata: 7 | labels: 8 | machineconfiguration.openshift.io/role: worker 9 | name: 100-intel-iommu-on 10 | spec: 11 | config: 12 | ignition: 13 | version: 3.2.0 14 | kernelArguments: 15 | - intel_iommu=on,sm_on modules_load=vfio-pci vfio-pci.ids=8086:4941,8086:4943 16 | selector: 17 | intel.feature.node.kubernetes.io/qat: 'true' 18 | intel.feature.node.kubernetes.io/dsa: 'true' 19 | -------------------------------------------------------------------------------- /tests/l2/dgpu/clinfo_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-dgpu-clinfo 8 | namespace: intel-dgpu 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: clinfo-pod 16 | image: image-registry.openshift-image-registry.svc:5000/intel-dgpu/intel-dgpu-clinfo:latest 17 | command: ["clinfo"] 18 | resources: 19 | limits: 20 | gpu.intel.com/i915: 1 21 | imagePullPolicy: IfNotPresent -------------------------------------------------------------------------------- /tests/l2/dgpu/intelvpl_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-dgpu-intelvpl 8 | namespace: intel-dgpu 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: intelvpl-pod 16 | image: image-registry.openshift-image-registry.svc:5000/intel-dgpu/intel-dgpu-intelvpl:latest 17 | command: ["vpl-inspect"] 18 | resources: 19 | limits: 20 | gpu.intel.com/i915: 1 21 | imagePullPolicy: IfNotPresent 22 | -------------------------------------------------------------------------------- /device_plugins/dsa_device_plugin.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | kind: DsaDevicePlugin 5 | apiVersion: deviceplugin.intel.com/v1 6 | metadata: 7 | name: dsadeviceplugin-sample 8 | spec: 9 | image: registry.connect.redhat.com/intel/intel-dsa-plugin@sha256:2742a13279cc3f301daa09b6389517024530f658d4e1dd13db495cc94d9ba57c 10 | initImage: registry.connect.redhat.com/intel/intel-idxd-config-initcontainer@sha256:b74dc43fa81ce14ea97f20ff6b2f726039f6309fdd868d5f45d751d0a8662cc1 11 | logLevel: 4 12 | nodeSelector: 13 | intel.feature.node.kubernetes.io/dsa: 'true' 14 | sharedDevNum: 10 15 | -------------------------------------------------------------------------------- /tests/gaudi/l2/hl-smi_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: hl-smi-workload 8 | namespace: gaudi-validation 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: hl-smi-workload 16 | image: vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26 17 | command: ["hl-smi"] 18 | resources: 19 | limits: 20 | habana.ai/gaudi: 8 21 | imagePullPolicy: IfNotPresent -------------------------------------------------------------------------------- /tests/l2/dgpu/hwinfo_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-dgpu-hwinfo 8 | namespace: intel-dgpu 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: hwinfo-pod 16 | image: image-registry.openshift-image-registry.svc:5000/intel-dgpu/intel-dgpu-hwinfo:latest 17 | command: ["hwinfo"] 18 | args: ["--display"] 19 | resources: 20 | limits: 21 | gpu.intel.com/i915: 1 22 | imagePullPolicy: IfNotPresent -------------------------------------------------------------------------------- /tests/l2/sgx/sgx_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-sgx-job 8 | namespace: intel-sgx 9 | spec: 10 | template: 11 | spec: 12 | restartPolicy: Never 13 | containers: 14 | - name: intel-sgx-job 15 | image: image-registry.openshift-image-registry.svc:5000/intel-sgx/intel-sgx-workload:latest 16 | imagePullPolicy: Always 17 | workingDir: "/opt/intel/" 18 | command: ["/opt/intel/app"] 19 | resources: 20 | limits: 21 | sgx.intel.com/epc: "5Mi" -------------------------------------------------------------------------------- /tests/gaudi/l2/hl-smi-firmware_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: hl-smi-firmware 8 | namespace: gaudi-validation 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: hl-smi-firmware 16 | image: vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26 17 | command: ["/bin/bash", "-c", "hl-smi -L | grep SPI"] 18 | resources: 19 | limits: 20 | habana.ai/gaudi: 1 21 | imagePullPolicy: IfNotPresent 22 | -------------------------------------------------------------------------------- /device_plugins/qat_device_plugin.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: deviceplugin.intel.com/v1 5 | kind: QatDevicePlugin 6 | metadata: 7 | name: qatdeviceplugin-sample 8 | spec: 9 | image: registry.connect.redhat.com/intel/intel-qat-plugin@sha256:8d79dba051b83ec770a4b0fdc3da6ac92264cb19cac8d455b707ed92a6a95d02 10 | initImage: registry.connect.redhat.com/intel/intel-qat-initcontainer@sha256:34f0b993ca654ea0b386217cba1a44d5ef3da841b3befc780508f5323e95fa90 11 | dpdkDriver: vfio-pci 12 | kernelVfDrivers: 13 | - 4xxxvf 14 | maxNumDevices: 128 15 | logLevel: 4 16 | nodeSelector: 17 | intel.feature.node.kubernetes.io/qat: "true" -------------------------------------------------------------------------------- /tests/l2/dgpu/vainfo_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-dgpu-vainfo 8 | namespace: intel-dgpu 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | containers: 15 | - name: vainfo-pod 16 | image: image-registry.openshift-image-registry.svc:5000/intel-dgpu/intel-dgpu-vainfo:latest 17 | command: ["vainfo"] 18 | args: ["--display", "drm", "--device", "/dev/dri/renderD128"] 19 | resources: 20 | limits: 21 | gpu.intel.com/i915: 1 22 | imagePullPolicy: IfNotPresent 23 | 24 | -------------------------------------------------------------------------------- /device_plugins/install_operator.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: operators.coreos.com/v1alpha2 5 | kind: OperatorGroup 6 | metadata: 7 | name: global-operators 8 | namespace: openshift-operators 9 | --- 10 | apiVersion: operators.coreos.com/v1alpha1 11 | kind: Subscription 12 | metadata: 13 | labels: 14 | operators.coreos.com/intel-device-plugins-operator.openshiftoperators: "" 15 | name: intel-device-plugins-operator 16 | namespace: openshift-operators 17 | spec: 18 | channel: alpha 19 | installPlanApproval: Automatic 20 | name: intel-device-plugins-operator 21 | source: certified-operators 22 | sourceNamespace: openshift-marketplace -------------------------------------------------------------------------------- /gaudi/gaudi_spi_fw_upgrade_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: gaudi-spi-firmware-upgrade 8 | namespace: gaudi-spi-fw-upgrade 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | serviceAccountName: gaudi-fw-upgrade-sa 15 | containers: 16 | - name: gaudi-spi-firmware-upgrade 17 | securityContext: 18 | privileged: true 19 | image: image-registry.openshift-image-registry.svc:5000/gaudi-spi-fw-upgrade/gaudi-spi-fw-upgrade:1.19.1-26 20 | command: [ "hl-fw-loader", "-y" ] 21 | imagePullPolicy: Always -------------------------------------------------------------------------------- /workloads/opea/chatqna/create_megaservice_container.sh: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | #!/bin/sh 4 | 5 | tag="v0.8" 6 | namespace="opea-chatqna" 7 | repo="https://github.com/opea-project/GenAIExamples.git" 8 | yaml_url="https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/workloads/opea/chatqna/chatqna_megaservice_buildconfig.yaml" 9 | 10 | oc project $namespace && 11 | git clone --depth 1 --branch $tag $repo && 12 | cd GenAIExamples/ChatQnA/deprecated/langchain/docker && 13 | oc extract secret/knative-serving-cert -n istio-system --to=. --keys=tls.crt && 14 | oc apply -f $yaml_url && 15 | oc start-build chatqna-megaservice --from-dir=./ --follow -------------------------------------------------------------------------------- /_templates/versions.html: -------------------------------------------------------------------------------- 1 | {%- if versions %} 2 |
3 | 4 | Versions 5 | {{ version }} 6 | 7 | 8 |
9 | {% if versions|length >= 1 %} 10 |
11 |
{{ _('Versions') }}
12 | {% for slug, url in versions %} 13 | {% if slug == version %} {% endif %} 14 |
{{ slug }}
15 | {% if slug == version %} {% endif %} 16 | {% endfor %} 17 |
18 | {% endif %} 19 |
20 |
21 | {%- endif %} -------------------------------------------------------------------------------- /security/qatlib_rbac.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: v1 5 | kind: ServiceAccount 6 | metadata: 7 | name: intel-qat 8 | namespace: intel-qat 9 | --- 10 | apiVersion: rbac.authorization.k8s.io/v1 11 | kind: Role 12 | metadata: 13 | name: intel-qat 14 | namespace: intel-qat 15 | rules: 16 | - apiGroups: 17 | - security.openshift.io 18 | resources: 19 | - securitycontextconstraints 20 | resourceNames: 21 | - intel-qat-scc 22 | verbs: 23 | - use 24 | --- 25 | apiVersion: rbac.authorization.k8s.io/v1 26 | kind: RoleBinding 27 | metadata: 28 | name: intel-qat 29 | roleRef: 30 | apiGroup: rbac.authorization.k8s.io 31 | kind: Role 32 | name: intel-qat 33 | subjects: 34 | - kind: ServiceAccount 35 | name: intel-qat 36 | namespace: intel-qat -------------------------------------------------------------------------------- /set-version.sh: -------------------------------------------------------------------------------- 1 | #!/bin/sh -eu 2 | # 3 | # Copyright 2024 Intel Corporation. 4 | # 5 | # SPDX-License-Identifier: Apache-2.0 6 | # 7 | # Invoke this script with a version as parameter 8 | # and it will update all hard-coded devel versions 9 | # to the tag versions in the source code. 10 | # 11 | # Adapted from https://github.com/intel/intel-device-plugins-for-kubernetes/ 12 | 13 | if [ $# != 1 ] || [ "$1" = "?" ] || [ "$1" = "--help" ]; then 14 | echo "Please provide TAG version as an argument. Usage: $0 " >&2 15 | exit 1 16 | fi 17 | 18 | devel_link="intel/intel-technology-enabling-for-openshift/main/" 19 | tag_link="intel/intel-technology-enabling-for-openshift/$1/" 20 | 21 | files=$(git grep -lF $devel_link -- '*.md') 22 | 23 | for file in $files; do 24 | sed -i -e "s|$devel_link|$tag_link|g" "$file"; 25 | done -------------------------------------------------------------------------------- /kmmo/intel-dgpu.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: kmm.sigs.x-k8s.io/v1beta1 5 | kind: Module 6 | metadata: 7 | name: intel-dgpu 8 | namespace: openshift-kmm 9 | spec: 10 | moduleLoader: 11 | container: 12 | modprobe: 13 | moduleName: i915 14 | firmwarePath: /firmware 15 | modulesLoadingOrder: 16 | - i915 17 | - mei_gsc 18 | inTreeModulesToRemove: [i915, intel_vsec, mei_gsc, mei_me] 19 | kernelMappings: 20 | - regexp: '^.*\.x86_64$' 21 | containerImage: registry.connect.redhat.com/intel/intel-data-center-gpu-driver-container:3.0.0-$KERNEL_FULL_VERSION 22 | selector: 23 | intel.feature.node.kubernetes.io/gpu: 'true' 24 | intel.feature.node.kubernetes.io/dgpu-canary: 'true' 25 | -------------------------------------------------------------------------------- /gaudi/gaudi_install_operator.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | # Adapted from https://docs.habana.ai/en/latest/Orchestration/HabanaAI_Operator/Deploying_HabanaAI_Operator.html#using-cli 4 | # 5 | --- 6 | apiVersion: v1 7 | kind: Namespace 8 | metadata: 9 | name: habana-ai-operator 10 | --- 11 | apiVersion: operators.coreos.com/v1 12 | kind: OperatorGroup 13 | metadata: 14 | name: habana-ai-operator 15 | namespace: habana-ai-operator 16 | spec: 17 | targetNamespaces: 18 | - habana-ai-operator 19 | --- 20 | apiVersion: operators.coreos.com/v1alpha1 21 | kind: Subscription 22 | metadata: 23 | name: habana-ai-operator 24 | namespace: habana-ai-operator 25 | spec: 26 | channel: stable 27 | installPlanApproval: Automatic 28 | name: habana-ai-operator 29 | source: certified-operators 30 | sourceNamespace: openshift-marketplace -------------------------------------------------------------------------------- /tests/l2/qat/qatlib_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-qat-workload 8 | namespace: intel-qat 9 | spec: 10 | template: 11 | spec: 12 | restartPolicy: Never 13 | containers: 14 | - name: intel-qat-job 15 | image: image-registry.openshift-image-registry.svc:5000/intel-qat/intel-qat-workload:latest 16 | imagePullPolicy: IfNotPresent 17 | command: ["cpa_sample_code"] 18 | securityContext: 19 | capabilities: 20 | add: 21 | [IPC_LOCK] 22 | resources: 23 | requests: 24 | qat.intel.com/dc: '1' 25 | qat.intel.com/cy: '1' 26 | limits: 27 | qat.intel.com/dc: '1' 28 | qat.intel.com/cy: '1' 29 | serviceAccount: intel-qat -------------------------------------------------------------------------------- /tests/l2/dsa/dsa_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: intel-dsa-workload 8 | namespace: intel-dsa 9 | spec: 10 | template: 11 | spec: 12 | restartPolicy: Never 13 | containers: 14 | - name: intel-dsa-job 15 | image: image-registry.openshift-image-registry.svc:5000/intel-dsa/intel-dsa-workload:latest 16 | imagePullPolicy: IfNotPresent 17 | workingDir: "/usr/libexec/accel-config/test/" 18 | command: 19 | - "./dsa_user_test_runner.sh" 20 | args: 21 | - "--skip-config" 22 | securityContext: 23 | capabilities: 24 | add: 25 | [SYS_RAWIO] 26 | resources: 27 | limits: 28 | dsa.intel.com/wq-user-dedicated: 1 29 | serviceAccountName: intel-dsa -------------------------------------------------------------------------------- /make.bat: -------------------------------------------------------------------------------- 1 | @ECHO OFF 2 | 3 | pushd %~dp0 4 | 5 | REM Command file for Sphinx documentation 6 | 7 | if "%SPHINXBUILD%" == "" ( 8 | set SPHINXBUILD=sphinx-build 9 | ) 10 | set SOURCEDIR=. 11 | set BUILDDIR=_build 12 | 13 | %SPHINXBUILD% >NUL 2>NUL 14 | if errorlevel 9009 ( 15 | echo. 16 | echo.The 'sphinx-build' command was not found. Make sure you have Sphinx 17 | echo.installed, then set the SPHINXBUILD environment variable to point 18 | echo.to the full path of the 'sphinx-build' executable. Alternatively you 19 | echo.may add the Sphinx directory to PATH. 20 | echo. 21 | echo.If you don't have Sphinx installed, grab it from 22 | echo.https://www.sphinx-doc.org/ 23 | exit /b 1 24 | ) 25 | 26 | if "%1" == "" goto help 27 | 28 | %SPHINXBUILD% -M %1 %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 29 | goto end 30 | 31 | :help 32 | %SPHINXBUILD% -M help %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 33 | 34 | :end 35 | popd 36 | -------------------------------------------------------------------------------- /playbooks/install_device_plugins_operator.yaml: -------------------------------------------------------------------------------- 1 | - name: Create global OperatorGroup in openshift-operators namespace 2 | k8s: 3 | state: present 4 | definition: 5 | apiVersion: operators.coreos.com/v1alpha2 6 | kind: OperatorGroup 7 | metadata: 8 | name: global-operators 9 | namespace: openshift-operators 10 | 11 | - name: Create Intel Device Plugins Operator Subscription 12 | k8s: 13 | state: present 14 | definition: 15 | apiVersion: operators.coreos.com/v1alpha1 16 | kind: Subscription 17 | metadata: 18 | labels: 19 | operators.coreos.com/intel-device-plugins-operator.openshiftoperators: "" 20 | name: intel-device-plugins-operator 21 | namespace: openshift-operators 22 | spec: 23 | channel: alpha 24 | installPlanApproval: Automatic 25 | name: intel-device-plugins-operator 26 | source: certified-operators 27 | sourceNamespace: openshift-marketplace -------------------------------------------------------------------------------- /tests/gaudi/l2/hccl_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: hccl-demo-workload 8 | namespace: gaudi-validation 9 | --- 10 | kind: BuildConfig 11 | apiVersion: build.openshift.io/v1 12 | metadata: 13 | name: hccl-demo-workload 14 | namespace: gaudi-validation 15 | spec: 16 | output: 17 | to: 18 | kind: ImageStreamTag 19 | name: 'hccl-demo-workload:1.19.1-26' 20 | strategy: 21 | type: Docker 22 | source: 23 | type: Dockerfile 24 | dockerfile: | 25 | ARG BUILDER=vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26 26 | FROM ${BUILDER} AS builder 27 | 28 | WORKDIR / 29 | RUN git clone https://github.com/HabanaAI/hccl_demo.git \ 30 | && cd hccl_demo \ 31 | && make 32 | 33 | WORKDIR /hccl_demo 34 | triggers: 35 | - type: ConfigChange 36 | runPolicy: Serial -------------------------------------------------------------------------------- /security/dsa_scc.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: security.openshift.io/v1 5 | allowHostDirVolumePlugin: false 6 | allowHostIPC: false 7 | allowHostNetwork: false 8 | allowHostPID: false 9 | allowHostPorts: false 10 | allowPrivilegeEscalation: false 11 | allowPrivilegedContainer: false 12 | allowedCapabilities: 13 | - SYS_RAWIO 14 | defaultAddCapabilities: null 15 | fsGroup: 16 | type: MustRunAs 17 | groups: [] 18 | kind: SecurityContextConstraints 19 | metadata: 20 | annotations: 21 | kubernetes.io/description: 'SCC for Intel DSA based workload' 22 | name: intel-dsa-scc 23 | priority: null 24 | readOnlyRootFilesystem: false 25 | requiredDropCapabilities: 26 | - ALL 27 | runAsUser: 28 | type: RunAsAny 29 | seLinuxContext: 30 | type: MustRunAs 31 | supplementalGroups: 32 | type: RunAsAny 33 | seccompProfiles: 34 | - runtime/default 35 | volumes: 36 | - configMap 37 | - downwardAPI 38 | - emptyDir 39 | - ephemeral 40 | - persistentVolumeClaim 41 | - projected 42 | - secret -------------------------------------------------------------------------------- /security/qatlib_scc.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: security.openshift.io/v1 5 | allowHostDirVolumePlugin: false 6 | allowHostIPC: false 7 | allowHostNetwork: false 8 | allowHostPID: false 9 | allowHostPorts: false 10 | allowPrivilegeEscalation: false 11 | allowPrivilegedContainer: false 12 | allowedCapabilities: 13 | - IPC_LOCK 14 | defaultAddCapabilities: null 15 | fsGroup: 16 | type: MustRunAs 17 | groups: [] 18 | kind: SecurityContextConstraints 19 | metadata: 20 | annotations: 21 | kubernetes.io/description: 'SCC for Intel QAT based workload' 22 | name: intel-qat-scc 23 | priority: null 24 | readOnlyRootFilesystem: false 25 | requiredDropCapabilities: 26 | - ALL 27 | runAsUser: 28 | type: RunAsAny 29 | seLinuxContext: 30 | type: MustRunAs 31 | supplementalGroups: 32 | type: RunAsAny 33 | seccompProfiles: 34 | - runtime/default 35 | volumes: 36 | - configMap 37 | - downwardAPI 38 | - emptyDir 39 | - ephemeral 40 | - persistentVolumeClaim 41 | - projected 42 | - secret -------------------------------------------------------------------------------- /index.rst: -------------------------------------------------------------------------------- 1 | .. Intel Technology Enabling for OpenShift documentation master file, created by 2 | sphinx-quickstart on Wed Apr 17 23:49:16 2024. 3 | You can adapt this file completely to your liking, but it should at least 4 | contain the root `toctree` directive. 5 | 6 | Intel® Technology Enabling for OpenShift* 7 | =================================================================== 8 | 9 | .. toctree:: 10 | :maxdepth: 2 11 | :caption: Contents: 12 | 13 | README.md 14 | nfd/README.md 15 | machine_configuration/README.md 16 | kmmo/README.md 17 | device_plugins/README.md 18 | gaudi/README.md 19 | 20 | .. toctree:: 21 | :maxdepth: 2 22 | :caption: One-Click Deployment: 23 | 24 | one_click/README.md 25 | 26 | .. toctree:: 27 | :maxdepth: 2 28 | :caption: End-to-end Solutions: 29 | 30 | e2e/inference/README.md 31 | 32 | .. toctree:: 33 | :maxdepth: 2 34 | :caption: Releases: 35 | 36 | docs/releases.rst 37 | 38 | .. toctree:: 39 | :maxdepth: 2 40 | :caption: Supported Platforms: 41 | 42 | docs/supported_platforms -------------------------------------------------------------------------------- /workloads/opea/chatqna/persistent_volumes.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: v1 5 | kind: PersistentVolume 6 | metadata: 7 | name: chatqna-megaservice-pv-0 8 | spec: 9 | capacity: 10 | storage: 100Mi 11 | accessModes: 12 | - ReadWriteOnce 13 | persistentVolumeReclaimPolicy: Retain 14 | nfs: 15 | server: x.x.x.x # nfs server 16 | path: /nfs # nfs path 17 | --- 18 | apiVersion: v1 19 | kind: PersistentVolume 20 | metadata: 21 | name: chatqna-megaservice-pv-1 22 | spec: 23 | capacity: 24 | storage: 100Mi 25 | accessModes: 26 | - ReadWriteOnce 27 | persistentVolumeReclaimPolicy: Retain 28 | nfs: 29 | server: x.x.x.x # nfs server 30 | path: /nfs # nfs path 31 | --- 32 | apiVersion: v1 33 | kind: PersistentVolume 34 | metadata: 35 | name: chatqna-megaservice-pv-2 36 | spec: 37 | capacity: 38 | storage: 100Mi 39 | accessModes: 40 | - ReadWriteOnce 41 | persistentVolumeReclaimPolicy: Retain 42 | nfs: 43 | server: x.x.x.x # nfs server 44 | path: /nfs # nfs path -------------------------------------------------------------------------------- /playbooks/README.md: -------------------------------------------------------------------------------- 1 | # Intel Technology Enabling Ansible Playbooks 2 | 3 | ## Overview 4 | This directory contains Ansible playbooks designed to automate the deployment and configuration of Intel technologies on Red Hat OpenShift clusters. These playbooks streamline the Intel feature provisioning and validation process on OpenShift environments. 5 | 6 | ## Prerequisites 7 | Before running the playbook, ensure the following prerequisites are met: 8 | - Provisioned RHOCP Cluster 9 | - Red Hat Enterprise Linux (RHEL) system with [Ansible](https://docs.ansible.com/ansible/2.9/installation_guide/intro_installation.html#installing-ansible-on-rhel-centos-or-fedora) installed and configured with a `kubeconfig` to connect to your RHOCP cluster. 10 | 11 | ## Run the Playbook 12 | 13 | To run the ansible playbook, clone this repository to your RHEL system. Navigate to the directory containing the playbook. 14 | 15 | ```bash 16 | git clone https://github.com/intel/intel-technology-enabling-for-openshift.git 17 | 18 | cd intel-technology-enabling-for-openshift/ 19 | 20 | ansible-playbook playbooks/intel_ocp_provisioning.yaml 21 | ``` -------------------------------------------------------------------------------- /playbooks/install_nfd_operator.yaml: -------------------------------------------------------------------------------- 1 | - name: Create namespace for Node Feature Discovery 2 | k8s: 3 | state: present 4 | definition: 5 | apiVersion: v1 6 | kind: Namespace 7 | metadata: 8 | name: openshift-nfd 9 | 10 | - name: Create operator group for Node Feature Discovery 11 | k8s: 12 | state: present 13 | definition: 14 | apiVersion: operators.coreos.com/v1 15 | kind: OperatorGroup 16 | metadata: 17 | generateName: openshift-nfd- 18 | name: openshift-nfd 19 | namespace: openshift-nfd 20 | spec: 21 | targetNamespaces: 22 | - openshift-nfd 23 | 24 | - name: Subscribe to Node Feature Discovery operator 25 | k8s: 26 | state: present 27 | definition: 28 | apiVersion: operators.coreos.com/v1alpha1 29 | kind: Subscription 30 | metadata: 31 | name: nfd 32 | namespace: openshift-nfd 33 | spec: 34 | channel: "stable" 35 | installPlanApproval: Automatic 36 | name: nfd 37 | source: redhat-operators 38 | sourceNamespace: openshift-marketplace -------------------------------------------------------------------------------- /tests/gaudi/l2/hccl_job.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2025 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: batch/v1 5 | kind: Job 6 | metadata: 7 | name: hccl-demo-workload 8 | namespace: gaudi-validation 9 | spec: 10 | template: 11 | metadata: 12 | spec: 13 | restartPolicy: Never 14 | serviceAccountName: hccl-demo-anyuid-sa 15 | containers: 16 | - name: hccl-demo-workload 17 | image: image-registry.openshift-image-registry.svc:5000/gaudi-validation/hccl-demo-workload:1.19.1-26 18 | workingDir: "/hccl_demo" 19 | command: ["/bin/bash", "-c", "--"] 20 | ## sleep for 20 seconds to avoid race condition 21 | args: 22 | - | 23 | sleep 20 24 | python3 run_hccl_demo.py --nranks 8 --node_id 0 --size 32m --test all_reduce --loop 1000 --ranks_per_node 8 25 | sleep 20 26 | env: 27 | - name: HCCL_COMM_ID 28 | value: '127.0.0.1:5555' 29 | resources: 30 | limits: 31 | habana.ai/gaudi: 8 32 | imagePullPolicy: IfNotPresent 33 | -------------------------------------------------------------------------------- /kmmo/intel-dgpu-on-premise-build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | labels: 8 | app: intel-dgpu-driver-container-kmmo 9 | name: intel-dgpu-driver-container-kmmo 10 | namespace: openshift-kmm 11 | spec: {} 12 | --- 13 | apiVersion: kmm.sigs.x-k8s.io/v1beta1 14 | kind: Module 15 | metadata: 16 | name: intel-dgpu-on-premise-build 17 | namespace: openshift-kmm 18 | spec: 19 | moduleLoader: 20 | container: 21 | imagePullPolicy: Always 22 | modprobe: 23 | moduleName: i915 24 | firmwarePath: /firmware 25 | inTreeModuleToRemove: intel_vsec 26 | kernelMappings: 27 | - regexp: '^.*\.x86_64$' 28 | containerImage: image-registry.openshift-image-registry.svc:5000/openshift-kmm/intel-dgpu-driver-container-kmmo:$KERNEL_FULL_VERSION 29 | build: 30 | dockerfileConfigMap: 31 | name: intel-dgpu-dockerfile-configmap 32 | selector: 33 | intel.feature.node.kubernetes.io/gpu: 'true' 34 | intel.feature.node.kubernetes.io/dgpu-canary: 'true' 35 | -------------------------------------------------------------------------------- /playbooks/configure_nfd.yaml: -------------------------------------------------------------------------------- 1 | - name: NFD - Create NFD discovery CR 2 | k8s: 3 | state: present 4 | definition: '{{ item }}' 5 | wait: yes 6 | wait_condition: 7 | type: Available 8 | status: 'False' 9 | with_items: '{{ lookup("url", "https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/nfd/node-feature-discovery-openshift.yaml", split_lines=False) | from_yaml_all | list }}' 10 | when: (item is not none) 11 | - name: NFD - Create NFD rules instance CR 12 | k8s: 13 | state: present 14 | definition: '{{ item }}' 15 | wait: yes 16 | with_items: '{{ lookup("url", "https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/nfd/node-feature-rules-openshift.yaml", split_lines=False) | from_yaml_all | list }}' 17 | when: (item is not none) -------------------------------------------------------------------------------- /tests/l2/dgpu/hwinfo_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-dgpu-hwinfo 8 | namespace: intel-dgpu 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: intel-dgpu-hwinfo 15 | namespace: intel-dgpu 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | dockerfile: | 24 | ARG BUILDER=registry.access.redhat.com/ubi9-minimal:latest 25 | FROM ${BUILDER} 26 | RUN microdnf -y update && \ 27 | rpm -ivh https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm && \ 28 | microdnf install -y hwinfo 29 | ENTRYPOINT ["hwinfo"] 30 | strategy: 31 | type: Docker 32 | noCache: true 33 | dockerStrategy: 34 | buildArgs: 35 | - name: "BUILDER" 36 | value: "registry.access.redhat.com/ubi9-minimal:latest" 37 | output: 38 | to: 39 | kind: ImageStreamTag 40 | name: intel-dgpu-hwinfo:latest -------------------------------------------------------------------------------- /device_plugins/deploy_sgx.md: -------------------------------------------------------------------------------- 1 | # Create Intel SGX Device Plugin CR 2 | 3 | ## Create a CR via web console 4 | 1. Go to **Operator** -> **Installed Operators**. 5 | 2. Open **Intel Device Plugins Operator**. 6 | 3. Navigate to tab **Intel Software Guard Extensions Device Plugin**. 7 | 4. Click **Create SgxDevicePlugin** -> set correct parameters -> Click **Create** 8 | 5. Optional: If you want to make any customizations, select YAML view and edit the details. When you are done, click **Create**. 9 | 10 | ## Verify via web console 11 | 1. Verify CR by checking the status of **Workloads** -> **DaemonSet** -> **intel-sgx-plugin**. 12 | 2. Now `SgxDevicePlugin` is created. 13 | 14 | ## Create CR via CLI 15 | Apply the CR yaml file: 16 | ``` 17 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/device_plugins/sgx_device_plugin.yaml 18 | ``` 19 | 20 | ## Verify via CLI 21 | Verify that the device plugin CR is ready: 22 | ``` 23 | $ oc get SgxDevicePlugin 24 | ``` 25 | Output: 26 | ``` 27 | NAME DESIRED READY NODE SELECTOR AGE 28 | sgxdeviceplugin-sample 1 1 {"intel.feature.node.kubernetes.io/sgx":"true"} 2m 29 | ``` -------------------------------------------------------------------------------- /Makefile: -------------------------------------------------------------------------------- 1 | # Minimal makefile for Sphinx documentation 2 | # 3 | 4 | # You can set these variables from the command line, and also 5 | # from the environment for the first two. 6 | SPHINXOPTS ?= 7 | SPHINXBUILD ?= sphinx-build 8 | SOURCEDIR = . 9 | BUILDDIR = _build 10 | 11 | # Put it first so that "make" without argument is like "make help". 12 | help: 13 | @$(SPHINXBUILD) -M help "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O) 14 | 15 | .PHONY: help Makefile 16 | 17 | # Catch-all target: route all unknown targets to Sphinx using the new 18 | # "make mode" option. $(O) is meant as a shortcut for $(SPHINXOPTS). 19 | # %: Makefile 20 | # @$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O) 21 | 22 | # Generate doc site under _build/html with Sphinx. 23 | vhtml: _work/venv/.stamp 24 | . _work/venv/bin/activate && \ 25 | $(SPHINXBUILD) -M html "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O) 26 | cp docs/index.html $(BUILDDIR)/html/index.html 27 | 28 | html: 29 | $(SPHINXBUILD) -M html "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O) 30 | cp docs/index.html $(BUILDDIR)/html/index.html 31 | 32 | clean-html: 33 | rm -rf $(BUILDDIR)/html 34 | 35 | # Set up a Python3 environment with the necessary tools for document creation. 36 | _work/venv/.stamp: ./requirements.txt 37 | rm -rf ${@D} 38 | python3 -m venv ${@D} 39 | . ${@D}/bin/activate && pip install wheel && pip install -r $< 40 | touch $@ -------------------------------------------------------------------------------- /tests/l2/sgx/README.md: -------------------------------------------------------------------------------- 1 | ### Verify Intel® Software Guard Extensions (Intel® SGX) Provisioning 2 | This [SampleEnclave](https://github.com/intel/linux-sgx/tree/master/SampleCode/SampleEnclave) application workload from the Intel SGX SDK runs an Intel SGX enclave utilizing the EPC resource from the Intel SGX provisioning. 3 | 4 | * Build the container image. 5 | ``` 6 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/sgx/sgx_imagestream.yaml 7 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/sgx/sgx_build.yaml 8 | ``` 9 | 10 | * Deploy and run the workload. 11 | ``` 12 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/sgx/sgx_job.yaml 13 | ``` 14 | 15 | * Check the results. 16 | ``` 17 | $ oc get pods -n intel-sgx 18 | intel-sgx-job-4tnh5 0/1 Completed 0 2m10s 19 | intel-sgx-workload-1-build 0/1 Completed 0 30s 20 | ``` 21 | ``` 22 | $ oc logs intel-sgx-job-4tnh5 -n intel-sgx 23 | Checksum(0x0x7fffac6f41e0, 100) = 0xfffd4143 24 | Info: executing thread synchronization, please wait... 25 | Info: SampleEnclave successfully returned. 26 | Enter a character before exit ... 27 | ``` 28 | ## See Also 29 | For Intel SGX demos on vanilla Kubernetes, refer to [link](https://github.com/intel/intel-device-plugins-for-kubernetes/tree/main/demo/sgx-sdk-demo) 30 | -------------------------------------------------------------------------------- /gaudi/gaudi_spi_fw_upgrade_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: gaudi-spi-fw-upgrade 8 | namespace: gaudi-spi-fw-upgrade 9 | --- 10 | kind: BuildConfig 11 | apiVersion: build.openshift.io/v1 12 | metadata: 13 | name: gaudi-spi-fw-upgrade 14 | namespace: gaudi-spi-fw-upgrade 15 | spec: 16 | output: 17 | to: 18 | kind: ImageStreamTag 19 | name: gaudi-spi-fw-upgrade:1.19.1-26 20 | strategy: 21 | type: Docker 22 | source: 23 | type: Dockerfile 24 | dockerfile: >+ 25 | ARG 26 | BUILDER=vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26 27 | 28 | FROM ${BUILDER} AS builder 29 | 30 | RUN echo "[habanalabs]" > /etc/yum.repos.d/habanalabs.repo && \ 31 | echo "name=Habana RH9 Linux repo" >> /etc/yum.repos.d/habanalabs.repo && \ 32 | echo "baseurl=https://vault.habana.ai/artifactory/rhel/9/9.4" >> /etc/yum.repos.d/habanalabs.repo && \ 33 | echo "gpgkey=https://vault.habana.ai/artifactory/api/v2/repositories/rhel/keyPairs/primary/public" >> /etc/yum.repos.d/habanalabs.repo && \ 34 | echo "gpgcheck=1" >> /etc/yum.repos.d/habanalabs.repo 35 | 36 | RUN yum makecache && dnf install -y habanalabs-firmware-odm 37 | triggers: 38 | - type: ConfigChange 39 | runPolicy: Serial -------------------------------------------------------------------------------- /tests/l2/README.md: -------------------------------------------------------------------------------- 1 | # Verifying Intel Hardware Feature Provisioning 2 | ## Introduction 3 | After provisioning Intel hardware features on RHOCP, the respective hardware resources are exposed to the RHOCP cluster. The workload containers can request these resources. The following sample workloads help verify if these resources can be used as expected. These sample workloads container images are built and packaged on-premises through [RHOCP BuildConfig](https://docs.openshift.com/container-platform/4.14/cicd/builds/understanding-buildconfigs.html) and pushed to the embedded repository through [RHOCP ImageStream](https://docs.openshift.com/container-platform/4.14/openshift_images/image-streams-manage.html). 4 | 5 | ## Prerequisites 6 | • Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster). 7 | 8 | • Provisioning Intel HW features on RHOCP. Follow steps [here](/README.md#provisioning-intel-hardware-features-on-rhocp) 9 | 10 | ### Verify Intel® Data Center GPU provisioning 11 | Please refer to Intel DGPU provisioning validation tests [here](dgpu/README.md) 12 | 13 | ### Verify Intel® Software Guard Extensions (Intel® SGX) Provisioning 14 | Please refer to Intel SGX provisioning validation tests [here](sgx/README.md) 15 | 16 | ### Verify Intel® QuickAssist Technology provisioning 17 | Please refer to Intel QAT provisioning validation tests [here](qat/README.md) 18 | 19 | ### Verify Intel® Data Streaming Accelerator provisioning 20 | Please refer to Intel DSA provisioning validation tests [here](dsa/README.md) -------------------------------------------------------------------------------- /workloads/opea/chatqna/tgi_gaudi_servingruntime.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: serving.kserve.io/v1alpha1 5 | kind: ServingRuntime 6 | metadata: 7 | name: tgi-gaudi-serving-runtime 8 | spec: 9 | containers: 10 | - name: kserve-container 11 | image: ghcr.io/huggingface/tgi-gaudi:1.2.1 12 | args: 13 | - --model-id 14 | - /mnt/models/ 15 | - --port=8080 16 | - --num-shard=3 #Number of GPU's 17 | - --sharded=true 18 | - --json-output 19 | env: #Add variables according to the chosen model 20 | - name: HF_HOME 21 | value: /tmp/hf_home 22 | - name: HF_OFFLINE 23 | value: "1" 24 | - name: TRANSFORMERS_OFFLINE 25 | value: "1" 26 | - name: HF_HUB_CACHE 27 | value: /mnt/models 28 | - name: HUGGING_FACE_HUB_TOKEN 29 | valueFrom: 30 | secretKeyRef: 31 | key: HUGGING_FACE_HUB_TOKEN 32 | name: hf-token 33 | resources: 34 | limits: 35 | cpu: "16" 36 | memory: 128Gi 37 | requests: 38 | cpu: "16" 39 | memory: 128Gi 40 | readinessProbe: 41 | exec: 42 | command: 43 | - curl 44 | - localhost:8080/health 45 | initialDelaySeconds: 500 46 | livenessProbe: 47 | exec: 48 | command: 49 | - curl 50 | - localhost:8080/health 51 | initialDelaySeconds: 500 52 | ports: 53 | - containerPort: 8080 54 | protocol: TCP 55 | multiModel: false 56 | supportedModelFormats: 57 | - autoSelect: true 58 | name: llm -------------------------------------------------------------------------------- /tests/l2/qat/qatlib_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-qat-workload 8 | namespace: intel-qat 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: intel-qat-workload 15 | namespace: intel-qat 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | dockerfile: | 24 | 25 | ARG BUILDER=registry.access.redhat.com/ubi9:latest 26 | FROM ${BUILDER} 27 | RUN subscription-manager register --username=${USERNAME} --password=${PASSWORD} && \ 28 | subscription-manager attach --auto && \ 29 | dnf repolist --disablerepo=* && \ 30 | subscription-manager repos --enable codeready-builder-for-rhel-9-x86_64-rpms && \ 31 | dnf -y update && \ 32 | dnf install -y qatlib qatlib-tests 33 | strategy: 34 | type: Docker 35 | noCache: true 36 | dockerStrategy: 37 | buildArgs: 38 | - name: "BUILDER" 39 | value: "registry.access.redhat.com/ubi9:latest" 40 | env: 41 | - name: "USERNAME" 42 | valueFrom: 43 | secretKeyRef: 44 | key: username 45 | name: rh-auth 46 | - name: "PASSWORD" 47 | valueFrom: 48 | secretKeyRef: 49 | key: password 50 | name: rh-auth 51 | 52 | output: 53 | to: 54 | kind: ImageStreamTag 55 | name: intel-qat-workload:latest -------------------------------------------------------------------------------- /workloads/opea/chatqna/chatqna_megaservice_buildconfig.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: chatqna-megaservice 8 | namespace: opea-chatqna 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: chatqna-megaservice 15 | namespace: opea-chatqna 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | dockerfile: | 23 | FROM langchain/langchain:latest 24 | 25 | RUN apt-get update -y && apt-get install -y --no-install-recommends --fix-missing \ 26 | libgl1-mesa-glx \ 27 | libjemalloc-dev 28 | 29 | RUN useradd -m -s /bin/bash user && \ 30 | mkdir -p /home/user && \ 31 | chown -R user /home/user/ 32 | 33 | USER user 34 | COPY requirements.txt /tmp/requirements.txt 35 | 36 | USER root 37 | COPY tls.crt /rhoai-ca/tls.crt 38 | RUN cat /rhoai-ca/tls.crt | tee -a '/usr/lib/ssl/cert.pem' 39 | 40 | USER user 41 | RUN pip install --no-cache-dir --upgrade pip && \ 42 | pip install --no-cache-dir -r /tmp/requirements.txt 43 | 44 | ENV PYTHONPATH=$PYTHONPATH:/ws:/home/user:/home/user/qna-app/app 45 | 46 | WORKDIR /home/user/qna-app 47 | COPY qna-app /home/user/qna-app 48 | 49 | ENTRYPOINT ["/usr/bin/sleep", "infinity"] 50 | triggers: 51 | - type: ConfigChange 52 | runPolicy: SerialLatestOnly 53 | strategy: 54 | type: Docker 55 | dockerStrategy: {} 56 | postCommit: {} 57 | output: 58 | to: 59 | kind: ImageStreamTag 60 | name: chatqna-megaservice:latest -------------------------------------------------------------------------------- /tests/README.md: -------------------------------------------------------------------------------- 1 | # Test Plan 2 | 3 | ```{admonition} Disclaimer 4 | Please note that this module is currently under development and may contain partially implemented features, therefore it is not supported in the current release. 5 | ``` 6 | 7 | ## Overview 8 | To ensure the whole stack is working as expected and track bugs, a layer based test architecture is needed on OCP. This plan consists of four layers. The first and second layers mentioned below, would be a part of the future automation testing framework on each OCP (x,y,z) release. 9 | 10 | ### L1 First Layer: Host OS and Driver Interfaces 11 | Layer 1 test cases should be executed before deploying [Intel Device Plugins Operator](/device_plugins/README.md) and after deploying OOT drivers like [Intel Data Center GPU Driver on OpenShift](/kmmo/README.md). It includes test cases : 12 | * to check existence of in-tree and OOT tree drivers 13 | * for SELinux and host OS security 14 | * check for devices on all nodes 15 | 16 | ### L2 Second Layer: Device Plugin Resource Provisioning 17 | L2 test cases are executed after deploying the [Intel Device Plugins Operator](/device_plugins/README.md). Refer to [readme](l2/README.md). It includes: 18 | * Pod's resource allocation and scheduling 19 | * Simple workloads 20 | * Boundary testing for the resources 21 | * Future plan for any failure analysis needed during automation. 22 | 23 | ### L3 Third Layer: E2E solution 24 | L3 test cases are executed after the specific device plugin related [e2e solution](/e2e) has been deployed. Please refer to [L3 test cases](l3/README.md) for detail. 25 | 26 | ### L4 Fourth Layer: Reference Workloads 27 | This layer includes the reference final application/usecase for the user. It integrates the whole stack and is custom for each Intel hardware feature and device plugin. This layer is yet to be added in upcoming releases. -------------------------------------------------------------------------------- /workloads/opea/chatqna/redis_deployment_service.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: v1 5 | kind: ServiceAccount 6 | metadata: 7 | name: opea-chatqna 8 | namespace: opea-chatqna 9 | --- 10 | apiVersion: rbac.authorization.k8s.io/v1 11 | kind: Role 12 | metadata: 13 | name: opea-chatqna 14 | namespace: opea-chatqna 15 | rules: 16 | - apiGroups: 17 | - security.openshift.io 18 | resources: 19 | - securitycontextconstraints 20 | resourceNames: 21 | - anyuid 22 | verbs: 23 | - use 24 | --- 25 | apiVersion: rbac.authorization.k8s.io/v1 26 | kind: RoleBinding 27 | metadata: 28 | name: opea-chatqna 29 | roleRef: 30 | apiGroup: rbac.authorization.k8s.io 31 | kind: Role 32 | name: opea-chatqna 33 | subjects: 34 | - kind: ServiceAccount 35 | name: opea-chatqna 36 | namespace: opea-chatqna 37 | --- 38 | # Redis Vector DB deployment 39 | apiVersion: apps/v1 40 | kind: Deployment 41 | metadata: 42 | name: redis-vector-db 43 | namespace: opea-chatqna 44 | spec: 45 | replicas: 1 46 | selector: 47 | matchLabels: 48 | app: redis-vector-db 49 | template: 50 | metadata: 51 | labels: 52 | app: redis-vector-db 53 | spec: 54 | serviceAccount: opea-chatqna 55 | containers: 56 | - name: redis-vector-db 57 | image: redis/redis-stack:7.2.0-v9 58 | ports: 59 | - containerPort: 6379 60 | - containerPort: 8001 61 | --- 62 | # Redis Vector DB Service 63 | apiVersion: v1 64 | kind: Service 65 | metadata: 66 | name: redis-vector-db 67 | namespace: opea-chatqna 68 | spec: 69 | type: ClusterIP 70 | selector: 71 | app: redis-vector-db 72 | ports: 73 | - name: redis-service 74 | port: 6379 75 | targetPort: 6379 76 | - name: redis-insight 77 | port: 8001 78 | targetPort: 8001 -------------------------------------------------------------------------------- /gaudi/Gaudi-SPI-Firmware-Upgrade.md: -------------------------------------------------------------------------------- 1 | # Upgrade Intel Gaudi SPI Firmware 2 | To upgrade Intel Gaudi SPI Firmware, follow below steps: 3 | 4 | **NOTE:** Currently this is only supported on Single Node OpenShift cluster. Multi node cluster support will be added in the future. 5 | 6 | ## Prerequisites 7 | - Make sure Gaudi drivers are unloaded. 8 | - On Red Hat OpenShift, delete existing ClusterPolicy Custom Resource. Verify output on the node using below command: 9 | ``` 10 | lsmod | grep habana 11 | ``` 12 | - Check the firmware version following the [firmware version check](https://github.com/intel/intel-technology-enabling-for-openshift/tree/main/tests/gaudi/l2#check-firmware-version-with-hl-smi). 13 | 14 | ## SPI Firmware Upgrade 15 | Build the container image with `habanalabs-firmware-odm` tool: 16 | ``` 17 | oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/gaudi/gaudi_spi_fw_upgrade_build.yaml 18 | ``` 19 | Create service account with required permissions: 20 | ``` 21 | oc create sa gaudi-fw-upgrade-sa -n gaudi-spi-fw-upgrade 22 | oc adm policy add-scc-to-user privileged -z gaudi-fw-upgrade-sa -n gaudi-spi-fw-upgrade 23 | ``` 24 | Deploy and execute the SPI firmware upgrade tool: 25 | ``` 26 | oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/gaudi/gaudi_spi_fw_upgrade_job.yaml 27 | ``` 28 | 29 | Verify Output: 30 | ``` 31 | oc get pods 32 | 33 | NAME READY STATUS RESTARTS AGE 34 | gaudi-spi-firmware-upgrade-ndmjp 0/1 Completed 0 10m 35 | ``` 36 | ``` 37 | oc logs gaudi-spi-firmware-upgrade-ndmjp 38 | . 39 | . 40 | #### 41 | #### Finished sending firmware: OK 42 | ``` 43 | Verify by following the [firmware version check](https://github.com/intel/intel-technology-enabling-for-openshift/tree/main/tests/gaudi/l2#check-firmware-version-with-hl-smi). -------------------------------------------------------------------------------- /tests/l2/dsa/dsa_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: build.openshift.io/v1 5 | kind: BuildConfig 6 | metadata: 7 | name: intel-dsa-workload 8 | namespace: intel-dsa 9 | spec: 10 | triggers: 11 | - type: "ConfigChange" 12 | - type: "ImageChange" 13 | runPolicy: "Serial" 14 | source: 15 | type: Dockerfile 16 | dockerfile: | 17 | 18 | ARG BUILDER=registry.access.redhat.com/ubi9:latest 19 | FROM ${BUILDER} 20 | RUN subscription-manager register --username=${USERNAME} --password=${PASSWORD} && \ 21 | subscription-manager attach --auto && \ 22 | dnf repolist --disablerepo=* && \ 23 | subscription-manager repos --enable rhel-9-for-x86_64-baseos-rpms --enable codeready-builder-for-rhel-9-x86_64-rpms && \ 24 | dnf -y update && \ 25 | dnf install -y gcc g++ make cmake autoconf automake libtool pkg-config \ 26 | git asciidoc xmlto libuuid-devel json-c-devel zlib-devel openssl-devel \ 27 | pciutils accel-config 28 | RUN git clone -b accel-config-v4.1.8 https://github.com/intel/idxd-config && \ 29 | cd idxd-config && ./autogen.sh && ./configure CFLAGS='-g -O2' --prefix=/usr \ 30 | --sysconfdir=/etc --libdir=/usr/lib64 --enable-test=yes && make && make install 31 | strategy: 32 | type: Docker 33 | noCache: true 34 | dockerStrategy: 35 | buildArgs: 36 | - name: "BUILDER" 37 | value: "registry.access.redhat.com/ubi9:latest" 38 | env: 39 | - name: "USERNAME" 40 | valueFrom: 41 | secretKeyRef: 42 | key: username 43 | name: rh-auth 44 | - name: "PASSWORD" 45 | valueFrom: 46 | secretKeyRef: 47 | key: password 48 | name: rh-auth 49 | output: 50 | to: 51 | kind: ImageStreamTag 52 | name: intel-dsa-workload:latest -------------------------------------------------------------------------------- /tests/l2/dgpu/clinfo_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-dgpu-clinfo 8 | namespace: intel-dgpu 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: intel-dgpu-clinfo 15 | namespace: intel-dgpu 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | dockerfile: | 24 | ARG BUILDER=registry.access.redhat.com/ubi9-minimal:latest 25 | FROM ${BUILDER} 26 | 27 | ARG OCL_ICD_VERSION=ocl-icd-2.2.13-4.el9.x86_64 28 | ARG CLINFO_VERSION=clinfo-3.0.21.02.21-4.el9.x86_64 29 | 30 | RUN microdnf install -y \ 31 | glibc \ 32 | yum-utils 33 | 34 | # install intel-opencl, ocl-icd and clinfo 35 | RUN dnf install -y 'dnf-command(config-manager)' && \ 36 | dnf config-manager --add-repo https://repositories.intel.com/gpu/rhel/9.0/lts/2350/unified/intel-gpu-9.0.repo && \ 37 | dnf install -y intel-opencl \ 38 | https://mirror.stream.centos.org/9-stream/AppStream/x86_64/os/Packages/$OCL_ICD_VERSION.rpm \ 39 | https://dl.fedoraproject.org/pub/epel/9/Everything/x86_64/Packages/c/$CLINFO_VERSION.rpm && \ 40 | dnf clean all && dnf autoremove && rm -rf /var/lib/dnf/lists/* && \ 41 | rm -rf /etc/yum.repos.d/intel-graphics.repo 42 | strategy: 43 | type: Docker 44 | noCache: true 45 | dockerStrategy: 46 | buildArgs: 47 | - name: "BUILDER" 48 | value: "registry.access.redhat.com/ubi9-minimal:latest" 49 | - name: "OCL_ICD_VERSION" 50 | value: "ocl-icd-2.2.13-4.el9.x86_64" 51 | - name: "CLINFO_VERSION" 52 | value: "clinfo-3.0.21.02.21-4.el9.x86_64" 53 | output: 54 | to: 55 | kind: ImageStreamTag 56 | name: intel-dgpu-clinfo:latest -------------------------------------------------------------------------------- /nfd/node-feature-rules-openshift.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2022 - 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: nfd.openshift.io/v1alpha1 5 | kind: NodeFeatureRule 6 | metadata: 7 | name: intel-dp-devices 8 | namespace: openshift-nfd 9 | spec: 10 | rules: 11 | - name: "intel.gpu" 12 | labels: 13 | "intel.feature.node.kubernetes.io/gpu": "true" 14 | matchFeatures: 15 | - feature: pci.device 16 | matchExpressions: 17 | vendor: {op: In, value: ["8086"]} 18 | class: {op: In, value: ["0300", "0380"]} 19 | 20 | - name: "intel.qat" 21 | labels: 22 | "intel.feature.node.kubernetes.io/qat": "true" 23 | matchFeatures: 24 | - feature: pci.device 25 | matchExpressions: 26 | vendor: {op: In, value: ["8086"]} 27 | device: {op: In, value: ["4940", "4942", "4944"]} 28 | class: {op: In, value: ["0b40"]} 29 | - feature: kernel.loadedmodule 30 | matchExpressions: 31 | intel_qat: {op: Exists} 32 | 33 | - name: "intel.sgx" 34 | labels: 35 | "intel.feature.node.kubernetes.io/sgx": "true" 36 | extendedResources: 37 | sgx.intel.com/epc: "@cpu.security.sgx.epc" 38 | matchFeatures: 39 | - feature: cpu.cpuid 40 | matchExpressions: 41 | SGX: {op: Exists} 42 | SGXLC: {op: Exists} 43 | - feature: cpu.security 44 | matchExpressions: 45 | sgx.enabled: {op: IsTrue} 46 | - feature: kernel.config 47 | matchExpressions: 48 | X86_SGX: {op: Exists} 49 | - name: "intel.dsa" 50 | labels: 51 | "intel.feature.node.kubernetes.io/dsa": "true" 52 | matchFeatures: 53 | - feature: pci.device 54 | matchExpressions: 55 | vendor: {op: In, value: ["8086"]} 56 | device: {op: In, value: ["0b25"]} 57 | class: {op: In, value: ["0880"]} 58 | - feature: kernel.loadedmodule 59 | matchExpressions: 60 | idxd: {op: Exists} 61 | -------------------------------------------------------------------------------- /device_plugins/deploy_gpu.md: -------------------------------------------------------------------------------- 1 | # Create Intel GPU Device Plugin CR 2 | 3 | ## Create CR via web console 4 | 1. Go to **Operator** -> **Installed Operators**. 5 | 2. Open **Intel Device Plugins Operator**. 6 | 3. Navigate to tab **Intel GPU Device Plugin**. 7 | 4. Click **Create GpuDevicePlugin** -> set correct parameters -> Click **Create**. 8 | 5. Optional: If you want to make any customizations, select YAML view and edit the details. Once you are done, click **Create**. 9 | 10 | ## Verify via web console 11 | 1. Verify CR by checking the status of **Workloads** -> **DaemonSet** -> **intel-gpu-plugin**. 12 | 2. Now `GpuDevicePlugin` is created. 13 | 14 | ## Create CR via CLI 15 | Apply the CR yaml file: 16 | ``` 17 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/device_plugins/gpu_device_plugin.yaml 18 | ``` 19 | 20 | ## Verify via CLI 21 | Verify that the device plugin CR is ready: 22 | ``` 23 | $ oc get GpuDevicePlugin 24 | ``` 25 | Output: 26 | ``` 27 | NAME DESIRED READY NODE SELECTOR AGE 28 | gpudeviceplugin-sample 1 1 {"intel.feature.node.kubernetes.io/gpu":"true"} 3m12s 29 | ``` 30 | 31 | # Using Intel Data Center GPU resource exclusively 32 | In this release, we only verified and support the single Intel GPU `i915` resource dedicated to the single workload pod. To achieve this, we set `sharedDevNum: 1` and `preferredAllocationPolicy: none` as default options. 33 | As the cluster administrator, use the [gpu_device_plugin.yaml](/device_plugins/gpu_device_plugin.yaml) provided from the previous section Create CR via CLI or use the default options from Create CR via web Console. 34 | As the application owner, when claiming the i915 resource, make sure the resource limits and requests are set as shown below: 35 | ``` 36 | spec: 37 | containers: 38 | - name: gpu-pod 39 | resources: 40 | limits: 41 | gpu.intel.com/i915: 1 42 | requests: 43 | gpu.intel.com/i915: 1 44 | ``` 45 | For more details, please refer to this [issue](https://github.com/intel/intel-device-plugins-for-kubernetes/issues/1408). -------------------------------------------------------------------------------- /tests/l2/dgpu/vainfo_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-dgpu-vainfo 8 | namespace: intel-dgpu 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: intel-dgpu-vainfo 15 | namespace: intel-dgpu 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | dockerfile: | 24 | ARG BUILDER=registry.access.redhat.com/ubi9:latest 25 | FROM ${BUILDER} 26 | RUN subscription-manager register --username=${USERNAME} --password=${PASSWORD} && \ 27 | subscription-manager attach --auto && \ 28 | subscription-manager repos --enable rhel-9-for-x86_64-appstream-rpms && \ 29 | dnf -y update && \ 30 | dnf install -y flex bison gcc gcc-c++ make autoconf libtool cmake git gdb \ 31 | libva libva-devel libdrm-devel 32 | RUN dnf install -y 'dnf-command(config-manager)' && \ 33 | dnf config-manager --add-repo \ 34 | https://repositories.intel.com/gpu/rhel/9.2/lts/2350/unified/intel-gpu-9.2.repo 35 | RUN dnf -y update && \ 36 | dnf install -y libva-utils intel-gmmlib-devel 37 | RUN git clone -b intel-media-23.4.3 --single-branch https://github.com/intel/media-driver.git && \ 38 | cd media-driver && mkdir media-driver build && cd build && \ 39 | cmake -D ENABLE_PRODUCTION_KMD=ON ../ && make -j $(nproc) && make install 40 | ENTRYPOINT ["/bin/sh"] 41 | strategy: 42 | type: Docker 43 | noCache: true 44 | dockerStrategy: 45 | buildArgs: 46 | - name: "BUILDER" 47 | value: "registry.access.redhat.com/ubi9:latest" 48 | env: 49 | - name: "USERNAME" 50 | valueFrom: 51 | secretKeyRef: 52 | key: username 53 | name: rh-auth 54 | - name: "PASSWORD" 55 | valueFrom: 56 | secretKeyRef: 57 | key: password 58 | name: rh-auth 59 | output: 60 | to: 61 | kind: ImageStreamTag 62 | name: intel-dgpu-vainfo:latest -------------------------------------------------------------------------------- /tests/gaudi/l2/vllm_deployment.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | kind: PersistentVolumeClaim 5 | apiVersion: v1 6 | metadata: 7 | name: vllm-workload-pvc 8 | namespace: gaudi-validation 9 | spec: 10 | accessModes: 11 | - ReadWriteOnce 12 | resources: 13 | requests: 14 | storage: 60Gi 15 | storageClassName: "" # Add your storage class 16 | volumeMode: Filesystem 17 | --- 18 | apiVersion: apps/v1 19 | kind: Deployment 20 | metadata: 21 | name: vllm-workload 22 | namespace: gaudi-validation 23 | labels: 24 | app: vllm-workload 25 | spec: 26 | replicas: 1 27 | selector: 28 | matchLabels: 29 | app: vllm-workload 30 | template: 31 | metadata: 32 | labels: 33 | app: vllm-workload 34 | spec: 35 | containers: 36 | - name: vllm-container 37 | image: image-registry.openshift-image-registry.svc:5000/gaudi-validation/vllm-workload:latest 38 | command: [ "/bin/bash", "-c", "--" ] 39 | args: ["vllm serve meta-llama/Llama-3.1-8B"] # Add the model 40 | ports: 41 | - containerPort: 8000 42 | resources: 43 | limits: 44 | habana.ai/gaudi: 1 45 | env: 46 | - name: HF_TOKEN 47 | valueFrom: 48 | secretKeyRef: 49 | name: hf-token 50 | key: hf-token 51 | - name: HF_HOME 52 | value: /home/vllm/.cache/huggingface 53 | - name: HF_HUB_OFFLINE 54 | value: "0" 55 | imagePullPolicy: Always 56 | volumeMounts: 57 | - name: hf-cache 58 | mountPath: /home/vllm/.cache 59 | - name: shm 60 | mountPath: /dev/shm 61 | volumes: 62 | - name: hf-cache 63 | persistentVolumeClaim: 64 | claimName: vllm-workload-pvc 65 | - name: shm 66 | emptyDir: 67 | medium: Memory 68 | sizeLimit: "2Gi" 69 | livenessProbe: 70 | httpGet: 71 | path: /health 72 | port: 8000 73 | initialDelaySeconds: 60 74 | periodSeconds: 10 75 | readinessProbe: 76 | httpGet: 77 | path: /health 78 | port: 8000 79 | initialDelaySeconds: 60 80 | periodSeconds: 5 81 | -------------------------------------------------------------------------------- /tests/l2/dgpu/intelvpl_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: intel-dgpu-intelvpl 8 | namespace: intel-dgpu 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: intel-dgpu-intelvpl 15 | namespace: intel-dgpu 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | dockerfile: | 24 | ARG BUILDER=registry.access.redhat.com/ubi9:latest 25 | FROM ${BUILDER} 26 | RUN subscription-manager register --username=${USERNAME} --password=${PASSWORD} && \ 27 | subscription-manager attach --auto && \ 28 | subscription-manager repos --enable rhel-9-for-x86_64-appstream-rpms && \ 29 | dnf -y update && \ 30 | dnf install -y flex bison gcc gcc-c++ make autoconf libtool cmake git gdb \ 31 | libva libva-devel libdrm libdrm-devel 32 | RUN dnf install -y 'dnf-command(config-manager)' && \ 33 | dnf config-manager --add-repo \ 34 | https://repositories.intel.com/gpu/rhel/9.2/lts/2350/unified/intel-gpu-9.2.repo 35 | RUN dnf -y update && \ 36 | dnf install -y libva-utils intel-gmmlib-devel libvpl2 libvpl-devel libvpl-tools \ 37 | libmfx 38 | RUN git clone -b intel-media-23.4.3 --single-branch https://github.com/intel/media-driver.git && \ 39 | cd media-driver && mkdir media-driver build && cd build && \ 40 | cmake -D ENABLE_PRODUCTION_KMD=ON ../ && make -j $(nproc) && make install 41 | # we need this for testing samples 42 | RUN git clone https://github.com/intel/libvpl.git 43 | ENTRYPOINT ["/bin/sh"] 44 | strategy: 45 | type: Docker 46 | noCache: true 47 | dockerStrategy: 48 | buildArgs: 49 | - name: "BUILDER" 50 | value: "registry.access.redhat.com/ubi9:latest" 51 | env: 52 | - name: "USERNAME" 53 | valueFrom: 54 | secretKeyRef: 55 | key: username 56 | name: rh-auth 57 | - name: "PASSWORD" 58 | valueFrom: 59 | secretKeyRef: 60 | key: password 61 | name: rh-auth 62 | output: 63 | to: 64 | kind: ImageStreamTag 65 | name: intel-dgpu-intelvpl:latest -------------------------------------------------------------------------------- /CONTRIBUTING.md: -------------------------------------------------------------------------------- 1 | # Contributing 2 | 3 | ### License 4 | 5 | Intel Technology Enabling For OpenShift project is licensed under the terms in [Apache-2.0](LICENSE.txt). By contributing to the project, you agree to the license and copyright terms therein and release your contribution under these terms. 6 | 7 | ### Sign your work 8 | 9 | Please use the sign-off line at the end of the patch. Your signature certifies that you wrote the patch or otherwise have the right to pass it on as an open-source patch. The rules are pretty simple: if you can certify 10 | the below (from [developercertificate.org](http://developercertificate.org/)): 11 | 12 | ``` 13 | Developer Certificate of Origin 14 | Version 1.1 15 | 16 | Copyright (C) 2004, 2006 The Linux Foundation and its contributors. 17 | 660 York Street, Suite 102, 18 | San Francisco, CA 94110 USA 19 | 20 | Everyone is permitted to copy and distribute verbatim copies of this 21 | license document, but changing it is not allowed. 22 | 23 | Developer's Certificate of Origin 1.1 24 | 25 | By making a contribution to this project, I certify that: 26 | 27 | (a) The contribution was created in whole or in part by me and I 28 | have the right to submit it under the open source license 29 | indicated in the file; or 30 | 31 | (b) The contribution is based upon previous work that, to the best 32 | of my knowledge, is covered under an appropriate open source 33 | license and I have the right under that license to submit that 34 | work with modifications, whether created in whole or in part 35 | by me, under the same open source license (unless I am 36 | permitted to submit under a different license), as indicated 37 | in the file; or 38 | 39 | (c) The contribution was provided directly to me by some other 40 | person who certified (a), (b) or (c) and I have not modified 41 | it. 42 | 43 | (d) I understand and agree that this project and the contribution 44 | are public and that a record of the contribution (including all 45 | personal information I submit with it, including my sign-off) is 46 | maintained indefinitely and may be redistributed consistent with 47 | this project or the open source license(s) involved. 48 | ``` 49 | 50 | Then you just add a line to every git commit message: 51 | 52 | Signed-off-by: Joe Smith 53 | 54 | Use your real name (sorry, no pseudonyms or anonymous contributions.) 55 | 56 | If you set your `user.name` and `user.email` git configs, you can sign your 57 | commit automatically with `git commit -s`. 58 | -------------------------------------------------------------------------------- /nfd/README.md: -------------------------------------------------------------------------------- 1 | # Setting up Node Feature Discovery 2 | [Node Feature Discovery (NFD) Operator](https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/specialized_hardware_and_driver_enablement/psap-node-feature-discovery-operator) manages the deployment and lifecycle of the NFD add-on to detect hardware features and system configuration, such as PCI cards, kernel, operating system version, etc. 3 | 4 | ## Prerequisites 5 | - Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster). 6 | 7 | ## Install NFD Operator 8 | Follow the guide below to install the NFD operator using CLI or web console. 9 | - [Install from the CLI](https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/specialized_hardware_and_driver_enablement/psap-node-feature-discovery-operator#install-operator-cli_psap-node-feature-discovery-operator) 10 | - [Install from the web console](https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/specialized_hardware_and_driver_enablement/psap-node-feature-discovery-operator#install-operator-web-console_psap-node-feature-discovery-operator) 11 | 12 | ## Configure NFD Operator 13 | Note: As RHOCP cluster administrator, you might need to merge the NFD operator config from the following Custom Resources (CRs) with other NFD operator configs that are already applied on your cluster. 14 | 15 | 1. Create `NodeFeatureDiscovery` CR instance. 16 | ``` 17 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/nfd/node-feature-discovery-openshift.yaml 18 | ``` 19 | 20 | 2. Create `NodeFeatureRule` CR instance. 21 | ``` 22 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/nfd/node-feature-rules-openshift.yaml 23 | ``` 24 | 25 | ## Verification 26 | Use the following command to get the node name 27 | ``` 28 | $ oc get nodes 29 | ``` 30 | Use the command shown below to verify whether the nodes are labeled properly by NFD: 31 | ``` 32 | $ oc describe node | grep intel.feature.node.kubernetes.io 33 | ``` 34 | Example output: 35 | ``` 36 | intel.feature.node.kubernetes.io/dgpu-canary=true 37 | intel.feature.node.kubernetes.io/gpu=true 38 | ``` 39 | 40 | ## Labels Table 41 | | Label | Intel hardware feature | 42 | | ----- | ---------------------- | 43 | | `intel.feature.node.kubernetes.io/gpu=true` | Intel® Data Center GPU Flex Series or Intel® Data Center GPU Max Series | 44 | | `intel.feature.node.kubernetes.io/sgx=true` | Intel® SGX | 45 | | `intel.feature.node.kubernetes.io/qat=true` | Intel® QAT | 46 | | `intel.feature.node.kubernetes.io/dsa=true` | Intel® DSA | 47 | 48 | ## See Also 49 | -------------------------------------------------------------------------------- /docs/releases.rst: -------------------------------------------------------------------------------- 1 | Release Information 2 | =================== 3 | .. list-table:: 4 | :align: left 5 | :widths: 15 10 10 10 10 10 10 10 10 10 10 10 10 10 6 | 7 | * - **Release** 8 | - `1.6.1 `_ 9 | - `1.6.0 `_ 10 | - `1.5.2 `_ 11 | - `1.5.1 `_ 12 | - `1.5.0 `_ 13 | - `1.4.0 `_ 14 | - `1.3.1 `_ 15 | - `1.3.0 `_ 16 | - `1.2.1 `_ 17 | - `1.2.0 `_ 18 | - `1.1.0 `_ 19 | - `1.0.1 `_ 20 | - `1.0.0 `_ 21 | * - **Red Hat OpenShift Version** 22 | - 4.18 23 | - 4.18 24 | - 4.17 25 | - 4.17 26 | - 4.17 27 | - 4.16 28 | - 4.14 29 | - 4.15 30 | - 4.14 31 | - 4.14 32 | - 4.13 33 | - 4.12 34 | - 4.12 35 | * - **Documentation** 36 | - `1.6.1 `_ 37 | - `1.6.0 `_ 38 | - `1.5.2 `_ 39 | - `1.5.1 `_ 40 | - `1.5.0 `_ 41 | - `1.4.0 `_ 42 | - `1.3.1 `_ 43 | - `1.3.0 `_ 44 | - NA 45 | - NA 46 | - NA 47 | - NA 48 | - NA 49 | 50 | **NOTE:** Release 1.3.1 supports OPEA and Gaudi -------------------------------------------------------------------------------- /machine_configuration/README.md: -------------------------------------------------------------------------------- 1 | # Setting up Machine Configuration 2 | 3 | ## Introduction 4 | Machine configuration operation is used to configure [Red Hat Enterprise Linux CoreOS (RHCOS)](https://docs.openshift.com/container-platform/4.14/architecture/architecture-rhcos.html) on each node in a RHOCP cluster. 5 | 6 | [Machine config operator](https://github.com/openshift/machine-config-operator) (MCO) is provided by Red Hat to manage the operating system and machine configuration. In this project through the MCO, cluster administrators can configure and update the kernel to provision Intel Hardware features on the worker nodes. 7 | 8 | MCO is one of the technologies used in this project to manage the machine configuration. In current OCP, MCO might reboot the node to enable the machine configuration. Since rebooting the node is undesirable, alternative machine configuration technologies are under investigation. For more details, see this [issue](https://github.com/intel/intel-technology-enabling-for-openshift/issues/34). 9 | 10 | The best approach is to work with the RHCOS team to push the RHCOS configuration as the default configuration for a RHOCP cluster on [Day 0](https://www.ibm.com/cloud/architecture/content/course/red-hat-openshift-container-platform-day-2-ops/). 11 | 12 | For some general configuration, we recommend you set it up while provisioning the cluster on [Day 1](https://www.ibm.com/cloud/architecture/content/course/red-hat-openshift-container-platform-day-2-ops/). 13 | 14 | If the configuration cannot be set as the default setting, we recommend using some operator to set the configuration on the fly without rebooting the node on [Day 2](https://www.ibm.com/cloud/architecture/content/course/red-hat-openshift-container-platform-day-2-ops/). 15 | 16 | Any contribution in this area is welcome. 17 | 18 | ## Prerequisites 19 | - Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster). 20 | - Setup node feature discovery (NFD). Follow steps [here](/nfd/README.md). 21 | 22 | ## Machine Configuration for Provisioning Intel® QAT and Intel® DSA 23 | 24 | * Turn on `intel_iommu,sm_on` kernel parameter and load `vfio_pci` at boot for QAT and DSA provisioning 25 | 26 | ``` 27 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/machine_configuration/100-intel-iommu-on.yaml 28 | ``` 29 | 30 | Note: This will reboot the worker nodes when changing the kernel parameter through MCO. 31 | 32 | ## Verification 33 | Navigate to the node terminal on the web console (Compute -> Nodes -> Select a node -> Terminal). Run the following commands in the terminal. 34 | ``` 35 | $ cat /proc/cmdline 36 | ``` 37 | Ensure that `intel_iommu=on,sm_on` is present. 38 | 39 | ``` 40 | $ chroot /host 41 | $ lsmod | grep vfio_pci 42 | ``` 43 | Ensure that `vfio_pci` driver is present. 44 | 45 | ## See Also 46 | - [Red Hat OpenShift Container Platform Day-2 operations](https://www.ibm.com/cloud/architecture/content/course/red-hat-openshift-container-platform-day-2-ops/) 47 | -------------------------------------------------------------------------------- /playbooks/intel_ocp_provisioning.yaml: -------------------------------------------------------------------------------- 1 | - hosts: localhost 2 | gather_facts: no 3 | vars: 4 | kubeconfig_path: "~/.kube/mojave-config" 5 | environment: 6 | KUBECONFIG: "{{ kubeconfig_path }}" 7 | vars_prompt: 8 | - name: "install_operators" 9 | prompt: "Do you want to install operators? 'Yes' to install NFD Operator and Intel Device Plugins Operator, or 'No' to skip" 10 | private: no 11 | - name: "validation_feature" 12 | prompt: "Which Intel feature do you want to validate? Enter 1 for Intel SGX, 2 for Intel QAT, 3 for Intel DSA, 4 for Intel GPU" 13 | private: no 14 | 15 | tasks: 16 | - name: Validate Inputs 17 | block: 18 | - name: Invalid Install Operators Input 19 | fail: 20 | msg: "Invalid input for Install Operators. Please enter a valid option for Install Operators (Yes/No)." 21 | when: install_operators not in ["Yes", "No"] 22 | - name: Invalid Validation Feature Input 23 | fail: 24 | msg: "Invalid input for validation feature. Please enter a valid option (1-4)." 25 | when: validation_feature not in ["1", "2", "3", "4"] 26 | 27 | - name: Install Operators 28 | block: 29 | - name: NFD - Install NFD Operator 30 | include_tasks: install_nfd_operator.yaml 31 | - name: IDPO - Install Intel Device Plugins Operator 32 | include_tasks: install_device_plugins_operator.yaml 33 | - name: NFD - Wait until the nfd-operator-controller Deployment is available 34 | k8s_info: 35 | kind: Deployment 36 | wait: yes 37 | name: nfd-controller-manager 38 | label_selectors: 39 | - operators.coreos.com/nfd.openshift-nfd 40 | - control-plane=controller-manager 41 | namespace: openshift-nfd 42 | wait_condition: 43 | type: Available 44 | status: 'True' 45 | - name: NFD - Configure NFD Operator 46 | include_tasks: configure_nfd.yaml 47 | - name: IDPO - Wait until the inteldeviceplugins-controller-manager Deployment is available 48 | k8s_info: 49 | kind: Deployment 50 | name: inteldeviceplugins-controller-manager 51 | namespace: openshift-operators 52 | wait: yes 53 | wait_condition: 54 | type: Available 55 | status: 'True' 56 | reason: MinimumReplicasAvailable 57 | when: install_operators == "Yes" 58 | 59 | - name: Skip Operator Installation 60 | debug: 61 | msg: "Skipping operator installation as per user input." 62 | when: install_operators == "No" 63 | 64 | - name: Validate Intel SGX 65 | include_tasks: validate_sgx.yaml 66 | when: validation_feature == "1" 67 | 68 | - name: Validate Intel QAT 69 | include_tasks: validate_gpu.yaml 70 | when: validation_feature == "2" 71 | 72 | - name: Validate Intel DSA 73 | include_tasks: validate_qat.yaml 74 | when: validation_feature == "3" 75 | 76 | - name: Validate Intel GPU 77 | include_tasks: validate_dsa.yaml 78 | when: validation_feature == "4" -------------------------------------------------------------------------------- /tests/l2/sgx/sgx_build.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2023 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: build.openshift.io/v1 5 | kind: BuildConfig 6 | metadata: 7 | name: intel-sgx-workload 8 | namespace: intel-sgx 9 | spec: 10 | triggers: 11 | - type: "ConfigChange" 12 | - type: "ImageChange" 13 | runPolicy: "Serial" 14 | source: 15 | type: Dockerfile 16 | dockerfile: | 17 | ARG BUILDER=registry.access.redhat.com/ubi9:latest 18 | ARG BASE=registry.access.redhat.com/ubi9-minimal:latest 19 | ARG LINUX_SGX_VERSION=2.26 20 | FROM ${BUILDER} AS builder 21 | 22 | ARG SGX_SDK=sgx_linux_x64_sdk_2.26.100.0.bin 23 | ARG LINUX_SGX_VERSION 24 | 25 | RUN dnf -y update && \ 26 | dnf -y install \ 27 | wget \ 28 | openssl \ 29 | gcc-c++ \ 30 | make \ 31 | protobuf-c && \ 32 | dnf clean all 33 | 34 | # SGX SDK installed in /opt/intel directory 35 | WORKDIR /opt/intel 36 | RUN wget https://download.01.org/intel-sgx/sgx-linux/$LINUX_SGX_VERSION/distro/rhel9.4-server/$SGX_SDK \ 37 | && chmod +x $SGX_SDK \ 38 | && echo "yes" | ./$SGX_SDK \ 39 | && rm $SGX_SDK 40 | 41 | RUN cd sgxsdk/SampleCode/SampleEnclave \ 42 | && . /opt/intel/sgxsdk/environment \ 43 | && make 44 | FROM ${BASE} 45 | ARG LINUX_SGX_VERSION 46 | RUN microdnf -y update && \ 47 | microdnf -y install \ 48 | wget \ 49 | tar \ 50 | gzip && \ 51 | microdnf clean all && rm -rf /var/cache/dnf 52 | 53 | # Download SGX PSW and install SGX runtime components to create SGX enclave 54 | WORKDIR /opt/intel 55 | RUN wget https://download.01.org/intel-sgx/sgx-linux/$LINUX_SGX_VERSION/distro/rhel9.4-server/sgx_rpm_local_repo.tgz \ 56 | && sha256sum sgx_rpm_local_repo.tgz \ 57 | && tar xvf sgx_rpm_local_repo.tgz \ 58 | && rm -rf sgx_rpm_local_repo.tgz 59 | 60 | RUN cd sgx_rpm_local_repo && rpm -i \ 61 | libsgx-headers-$LINUX_SGX_VERSION* \ 62 | libsgx-enclave-common-$LINUX_SGX_VERSION* \ 63 | libsgx-urts-$LINUX_SGX_VERSION* && \ 64 | rm -r /opt/intel/sgx_rpm_local_repo 65 | 66 | COPY --from=builder /opt/intel/sgxsdk/SampleCode/SampleEnclave/app app 67 | COPY --from=builder /opt/intel/sgxsdk/SampleCode/SampleEnclave/enclave.signed.so enclave.signed.so 68 | 69 | ENTRYPOINT /opt/intel/app 70 | 71 | strategy: 72 | type: Docker 73 | noCache: true 74 | dockerStrategy: 75 | buildArgs: 76 | - name: "BUILDER" 77 | value: "registry.access.redhat.com/ubi9:9.4" 78 | - name: "BASE" 79 | value: "registry.access.redhat.com/ubi9-minimal:9.4" 80 | - name: "SGX_SDK" 81 | value: "sgx_linux_x64_sdk_2.26.100.0.bin" 82 | - name: "LINUX_SGX_VERSION" 83 | value: "2.26" 84 | output: 85 | to: 86 | kind: ImageStreamTag 87 | name: intel-sgx-workload:latest -------------------------------------------------------------------------------- /device_plugins/deploy_dsa.md: -------------------------------------------------------------------------------- 1 | # Create Intel DSA Device Plugin CR 2 | 3 | ## Create a CR via web console 4 | 1. Go to **Operator** -> **Installed Operators**. 5 | 2. Open **Intel Device Plugins Operator**. 6 | 3. Navigate to tab **Intel DSA Device Plugin**. 7 | 4. Click **Create DSADevicePlugin** -> set correct parameters -> Click **Create** 8 | 5. Optional: If you want to make any customizations, select YAML view and edit the details. When you are done, click **Create**. 9 | 10 | ## Verify via web console 11 | 1. Verify CR by checking the status of **Workloads** -> **DaemonSet** -> **intel-dsa-plugin**. 12 | 2. Now `DsaDevicePlugin` is created. 13 | 14 | ## Create CR via CLI 15 | Apply the CR yaml file: 16 | ``` 17 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/device_plugins/dsa_device_plugin.yaml 18 | ``` 19 | 20 | ## Verify via CLI 21 | Verify that the device plugin CR is ready: 22 | ``` 23 | $ oc get DsaDevicePlugin 24 | ``` 25 | Output: 26 | ``` 27 | NAME DESIRED READY NODE SELECTOR AGE 28 | dsadeviceplugin-sample 3 3 {"intel.feature.node.kubernetes.io/dsa":"true"} 98m 29 | ``` 30 | 31 | # Verify DSA Device Plugin 32 | After the plugin is deployed, use below command to verify DSA resources: 33 | ``` 34 | $ oc describe node srf-2 | grep dsa.intel.com 35 | dsa.intel.com/wq-user-dedicated: 0 36 | dsa.intel.com/wq-user-shared: 160 37 | dsa.intel.com/wq-user-dedicated: 0 38 | dsa.intel.com/wq-user-shared: 160 39 | ``` 40 | 41 | ## DSA Resource Configuration 42 | By default the DSA plugin uses [this configuration file](https://github.com/intel/intel-device-plugins-for-kubernetes/blob/main/demo/dsa.conf). 43 | The dsa init container comes with a utility called `accel-config` which takes file as input and configures the DSA hardwares based on that. 44 | The default configuration has creates dedicated WQs for each DSA device so that it's four groups per device where each groups is with 1 WQ linked to 1 engine. 45 | Users can customise the config and can use the pre-customised config for their specific use case from [here](https://github.com/intel/idxd-config/tree/stable/contrib/configs) 46 | There's also a possibility for a node specific configuration by passing a node specific profile via configMap volume mount. 47 | Users can use the steps below to customize the DSA resource configuration: 48 | 1. Create the configmap for DSA resource configuration 49 | ``` 50 | $ oc create configmap --namespace=openshift-operators intel-dsa-config --from-file=dsa[-$NODE_NAME].conf 51 | 2. Create DSA device plugin CR with -provisioning-config set as the name of the ConfigMap (created in step 1) in the dsa_device_plugin.yaml file or set ConfigMap name in the provisioning-config option from web console. 52 | 53 | # Run Intel DSA based workloads on RHOCP 54 | To run the Intel DSA based workloads as an unprivileged pod, you need to use a customised SCC. The customized `dsa-scc` Security Context Constraint (SCC) is provided to bind with service account and run the DSA based workload. 55 | 56 | See [Verify Intel DSA Provisioning](/tests/l2/dsa/README.md) for the detailed steps. 57 | -------------------------------------------------------------------------------- /tests/l2/qat/README.md: -------------------------------------------------------------------------------- 1 | ### Verify Intel® QuickAssist Technology provisioning 2 | This workload runs [qatlib](https://github.com/intel/qatlib) sample tests using RedHat built and distributed Qatlib RPM packages from the codeready-builder-for-rhel-9-x86_64-rpms repo. Refer to the [qatlib readme](https://github.com/intel/qatlib/blob/main/INSTALL) for more details. 3 | 4 | * Create and use ```intel-qat``` namespace for the workload 5 | 6 | ``` 7 | $ oc new-project intel-qat 8 | ``` 9 | 10 | * Build the workload container image 11 | 12 | Please replace the credentials in buildconfig yaml with your RedHat account login credentials. 13 | 14 | ``` 15 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/qat/qatlib_build.yaml 16 | ``` 17 | 18 | * Create SCC intel-qat-scc for Intel QAT based workload, if this SCC is not created 19 | 20 | ``` 21 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/qatlib_scc.yaml 22 | ``` 23 | 24 | * Create the intel-qat service account to use intel-qat-scc 25 | 26 | ``` 27 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/qatlib_rbac.yaml 28 | ``` 29 | 30 | * Deploy the qatlib workload job with intel-qat service account 31 | 32 | ``` 33 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/qat/qatlib_job.yaml 34 | ``` 35 | 36 | * Check the results. 37 | ``` 38 | $ oc get pods -n intel-qat 39 | intel-qat-workload-c6g9v 0/1 Completed 0 4m13s 40 | ``` 41 | 42 | 43 | * For all sample tests `cpa_sample_code` 44 | 45 | ``` 46 | $ oc logs intel-qat-workload-c6g9v -n intel-qat 47 | qaeMemInit started 48 | icp_sal_userStartMultiProcess("SSL") started 49 | There are no crypto instances 50 | *** QA version information *** 51 | device ID = 0 52 | software = 23.2.0 53 | *** END QA version information *** 54 | Inst 0, Affin: 0, Dev: 0, Accel 0, EE 0, BDF ED:00:01 55 | Inst 1, Affin: 1, Dev: 0, Accel 0, EE 0, BDF ED:00:01 56 | Inst 2, Affin: 2, Dev: 0, Accel 0, EE 0, BDF ED:00:01 57 | Inst 3, Affin: 3, Dev: 0, Accel 0, EE 0, BDF ED:00:01 58 | --------------------------------------- 59 | API Traditional 60 | Session State STATELESS 61 | Algorithm DEFLATE 62 | Huffman Type STATIC 63 | Mode ASYNCHRONOUS 64 | CNV Enabled YES 65 | Direction COMPRESS 66 | Packet Size 8192 67 | Compression Level 1 68 | Corpus CALGARY_CORPUS 69 | Corpus Filename calgary 70 | CNV Recovery Enabled YES 71 | Number of threads 4 72 | Total Responses 158400 73 | Total Retries 2242671 74 | Clock Cycles Start 126150916653843 75 | Clock Cycles End 126151409143747 76 | Total Cycles 492489904 77 | CPU Frequency(kHz) 1700160 78 | Throughput(Mbps) 35920 79 | Compression Ratio 0.4897 80 | --------------------------------------- 81 | 82 | Inst 0, Affin: 0, Dev: 0, Accel 0, EE 0, BDF ED:00:01 83 | Inst 1, Affin: 1, Dev: 0, Accel 0, EE 0, BDF ED:00:01 84 | Inst 2, Affin: 2, Dev: 0, Accel 0, EE 0, BDF ED:00:01 85 | Inst 3, Affin: 3, Dev: 0, Accel 0, EE 0, BDF ED:00:01 86 | --------------------------------------- 87 | ``` 88 | -------------------------------------------------------------------------------- /one_click/README.md: -------------------------------------------------------------------------------- 1 | # Deploy Intel Technology Enabling Solutions with Red Hat OpenShift using “One-Click” 2 | 3 | ## Overview 4 | Red Hat [Ansible](https://www.ansible.com/) and Operator technologies are used for “One-Click Deployment” of Intel technology enabling solutions with Red Hat OpenShift Container Platform (RHOCP). Ansible technology automates the operator installation and configuration steps using a playbook, making deployment as simple as a single click. 5 | 6 | The referenced Ansible playbooks here can be used by the cluster administrators to customize their own playbooks. 7 | 8 | **Note:** It is recommended to start from [Get started](/README.md#getting-started) to get familiar with the installation and configuration of the general operator before composing the first playbook. 9 | 10 | ## Reference Playbook – Intel Data Center GPU Provisioning 11 | 12 | This playbook demonstrates the one-click provisioning of Intel Data Center GPU on an RHOCP cluster. The steps involved are installation and configuration of general Operators including Node Feature Discovery (NFD) operator, Kernel Module Management (KMM) operator, and the Intel Device Plugins Operator. 13 | 14 | ### Prerequisite 15 | Before running the playbook, ensure the following prerequisites are met: 16 | - Provisioned RHOCP Cluster 17 | - Red Hat Enterprise Linux (RHEL) system with [Ansible](https://docs.ansible.com/ansible/2.9/installation_guide/intro_installation.html#installing-ansible-on-rhel-centos-or-fedora) installed and configured with a `kubeconfig` to connect to your RHOCP cluster. 18 | 19 | ### Run the Playbook 20 | To run the ansible playbook, clone this repository to your RHEL system. Navigate to the directory containing the playbook. 21 | ``` 22 | $ git clone https://github.com/intel/intel-technology-enabling-for-openshift.git 23 | $ cd intel-technology-enabling-for-openshift/one_click 24 | ``` 25 | Execute below single command to provision Intel Data Center GPU: 26 | ``` 27 | $ ansible-playbook gpu_provisioning_playbook.yaml 28 | ``` 29 | 30 | ## Reference Playbook – Intel Gaudi Provisioning 31 | This playbook demonstrates the one-click provisioning of Intel Gaudi AI Accelerator on an RHOCP cluster. The steps involved are installation and configuration of general Operators including Node Feature Discovery (NFD) operator, Kernel Module Management (KMM) operator, and the Intel Gaudi Base Operator. The playbook also creates the Gaudi `DeviceConfig` CR which deploys the Gaudi Out-of-Tree drivers, Gaudi device plugins, Habana container runtime and Habana node metrics. 32 | 33 | ### Prerequisite 34 | Before running the playbook, ensure the following prerequisites are met: 35 | - Provisioned RHOCP Cluster 36 | - Red Hat Enterprise Linux (RHEL) system with [Ansible](https://docs.ansible.com/ansible/2.9/installation_guide/intro_installation.html#installing-ansible-on-rhel-centos-or-fedora) installed and configured with a `kubeconfig` to connect to your RHOCP cluster. 37 | - Set Firmware search path using MCO, follow [Update Kernel Firmware Search Path with MCO](/gaudi/README.md#update-kernel-firmware-search-path-with-mco). 38 | 39 | ### Run the Playbook 40 | To run the ansible playbook, clone this repository to your RHEL system. Navigate to the directory containing the playbook. 41 | ``` 42 | $ git clone https://github.com/intel/intel-technology-enabling-for-openshift.git 43 | $ cd intel-technology-enabling-for-openshift/one_click 44 | ``` 45 | Execute below single command to provision Intel Gaudi Accelerator: 46 | ``` 47 | $ ansible-playbook gaudi_provisioning_playbook.yaml 48 | ``` -------------------------------------------------------------------------------- /workloads/opea/chatqna/chatqna_megaservice_deployment.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | --- 4 | apiVersion: v1 5 | kind: PersistentVolumeClaim 6 | metadata: 7 | name: chatqna-megaservice-pvc-0 8 | namespace: opea-chatqna 9 | spec: 10 | accessModes: 11 | - ReadWriteOnce 12 | resources: 13 | requests: 14 | storage: 100Mi 15 | --- 16 | apiVersion: v1 17 | kind: PersistentVolumeClaim 18 | metadata: 19 | name: chatqna-megaservice-pvc-1 20 | namespace: opea-chatqna 21 | spec: 22 | accessModes: 23 | - ReadWriteOnce 24 | resources: 25 | requests: 26 | storage: 100Mi 27 | --- 28 | apiVersion: v1 29 | kind: PersistentVolumeClaim 30 | metadata: 31 | name: chatqna-megaservice-pvc-2 32 | namespace: opea-chatqna 33 | spec: 34 | accessModes: 35 | - ReadWriteOnce 36 | resources: 37 | requests: 38 | storage: 100Mi 39 | --- 40 | apiVersion: apps/v1 41 | kind: Deployment 42 | metadata: 43 | name: chatqna-megaservice 44 | namespace: opea-chatqna 45 | spec: 46 | selector: 47 | matchLabels: 48 | app: chatqna-megaservice 49 | replicas: 1 50 | template: 51 | metadata: 52 | labels: 53 | app: chatqna-megaservice 54 | spec: 55 | serviceAccount: opea-chatqna 56 | containers: 57 | - name: chatqna-megaservice 58 | image: 'image-registry.openshift-image-registry.svc:5000/opea-chatqna/chatqna-megaservice:latest' 59 | env: 60 | - name: EMBED_MODEL 61 | value: BAAI/bge-base-en-v1.5 62 | - name: HUGGINGFACEHUB_API_TOKEN 63 | valueFrom: 64 | secretKeyRef: 65 | key: HUGGINGFACEHUB_API_TOKEN 66 | name: hf-token 67 | - name: MODEL_SIZE 68 | value: 70b 69 | - name: PYTHONPATH 70 | value: $PYTHONPATH:/ws:/home/user:/home/user/qna-app/app 71 | - name: RAG_UPLOAD_DIR 72 | value: /upload_dir 73 | - name: REDIS_PORT 74 | value: "6379" 75 | - name: REDIS_HOST 76 | value: "redis-vector-db" 77 | - name: REDIS_SCHEMA 78 | value: schema_dim_768.yml 79 | - name: TGI_ENDPOINT 80 | value: http://xxx.xxx.xxx.xxx:xxx 81 | ports: 82 | - containerPort: 8000 83 | command: 84 | - /bin/bash 85 | - '-c' 86 | - | 87 | cd /ws && \ 88 | python ingest.py /ws/data_intel/ && \ 89 | cd /home/user/qna-app && \ 90 | python app/server.py 91 | volumeMounts: 92 | - mountPath: /ws 93 | name: chatqna-megaservice-pvc-0 94 | - mountPath: /test 95 | name: chatqna-megaservice-pvc-1 96 | - mountPath: /upload_dir 97 | name: chatqna-megaservice-pvc-2 98 | volumes: 99 | - name: chatqna-megaservice-pvc-0 100 | persistentVolumeClaim: 101 | claimName: chatqna-megaservice-pvc-0 102 | - name: chatqna-megaservice-pvc-1 103 | persistentVolumeClaim: 104 | claimName: chatqna-megaservice-pvc-1 105 | - name: chatqna-megaservice-pvc-2 106 | persistentVolumeClaim: 107 | claimName: chatqna-megaservice-pvc-2 108 | --- 109 | # Chatqna megaservice Service 110 | apiVersion: v1 111 | kind: Service 112 | metadata: 113 | name: chatqna-megaservice 114 | namespace: opea-chatqna 115 | spec: 116 | type: ClusterIP 117 | selector: 118 | app: chatqna-megaservice 119 | ports: 120 | - port: 8000 121 | targetPort: 8000 -------------------------------------------------------------------------------- /tests/l2/dsa/README.md: -------------------------------------------------------------------------------- 1 | ### Verify Intel® Data Streaming Accelerator (DSA) Technology provisioning 2 | This workload runs [accel-config](https://github.com/intel/idxd-config) sample tests using RedHat built and distributed accel-config RPM packages from the rhel-9-for-x86_64-baseos-rpms repo. Refer to the [accel config readme](https://github.com/intel/idxd-config/blob/stable/README.md) for more details. 3 | 4 | * Create the RedHat auth secret. Please replace the credentials in the secret yaml with your RedHat account login credentials. 5 | 6 | ``` 7 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/dsa/rh_auth.yaml 8 | ``` 9 | 10 | * Build the workload container image 11 | 12 | ``` 13 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/dsa/dsa_imagestream.yaml 14 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/dsa/dsa_build.yaml 15 | ``` 16 | 17 | * Create SCC intel-dsa-scc for Intel DSA based workload, if this SCC is not created 18 | 19 | ``` 20 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/dsa_scc.yaml 21 | ``` 22 | 23 | * Create the intel-dsa serviceAccount, role and roleBinding to use intel-dsa-scc 24 | 25 | ``` 26 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/dsa_serviceAccount.yaml 27 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/dsa_role.yaml 28 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/security/dsa_roleBinding.yaml 29 | ``` 30 | 31 | * Deploy the accel-config workload job with intel-dsa service account 32 | 33 | ``` 34 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/tests/l2/dsa/dsa_job.yaml 35 | ``` 36 | 37 | * Check the results. 38 | ``` 39 | $ oc get pods -n intel-dsa 40 | intel-dsa-workload-244xm 0/1 Completed 0 3m12s 41 | ``` 42 | 43 | * sample test logs 44 | ``` 45 | $ oc logs intel-dsa-workload-244xm -n intel-dsa 46 | dsa0/wq0.1 47 | dsa0 48 | Testing with 'block on fault' flag ON 49 | Performing dedicated WQ NOOP testing 50 | Testing 1 bytes 51 | [ info] alloc wq 1 dedicated size 16 addr 0x7f0cde00b000 batch sz 0x400 xfer sz 0x80000000 52 | [ info] testnoop: tflags 0x1 num_desc 1 53 | [ info] preparing descriptor for noop 54 | [ info] Submitted all noop jobs 55 | [ info] verifying task result for 0x2041620 56 | [ info] test with op 0 passed 57 | Testing 4096 bytes 58 | [ info] alloc wq 1 dedicated size 16 addr 0x7fd4881da000 batch sz 0x400 xfer sz 0x80000000 59 | [ info] testnoop: tflags 0x1 num_desc 1 60 | [ info] preparing descriptor for noop 61 | [ info] Submitted all noop jobs 62 | [ info] verifying task result for 0x82f620 63 | [ info] test with op 0 passed 64 | Testing 65536 bytes 65 | [ info] alloc wq 1 dedicated size 16 addr 0x7f462bbed000 batch sz 0x400 xfer sz 0x80000000 66 | [ info] testnoop: tflags 0x1 num_desc 1 67 | [ info] preparing descriptor for noop 68 | [ info] Submitted all noop jobs 69 | [ info] verifying task result for 0xe4e620 70 | [ info] test with op 0 passed 71 | Testing 1048576 bytes 72 | [ info] alloc wq 1 dedicated size 16 addr 0x7fac2ac0c000 batch sz 0x400 xfer sz 0x80000000 73 | [ info] testnoop: tflags 0x1 num_desc 1 74 | [ info] preparing descriptor for noop 75 | [ info] Submitted all noop jobs 76 | [ info] verifying task result for 0xf21620 77 | [ info] test with op 0 passed 78 | Testing 2097152 bytes 79 | [ info] alloc wq 1 dedicated size 16 addr 0x7f7426a5c000 batch sz 0x400 xfer sz 0x80000000 80 | [ info] testnoop: tflags 0x1 num_desc 1 81 | [ info] preparing descriptor for noop 82 | [ info] Submitted all noop jobs 83 | [ info] verifying task result for 0xeec620 84 | [ info] test with op 0 passed 85 | Performing shared WQ NOOP testing 86 | ``` 87 | -------------------------------------------------------------------------------- /gaudi/gaudi_cluster_policy.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | # Adapted from https://docs.habana.ai/en/latest/Installation_Guide/Additional_Installation/Kubernetes_Installation/Kubernetes_Operator.html#id2 4 | # 5 | apiVersion: habanalabs.habana.ai/v1 6 | kind: ClusterPolicy 7 | metadata: 8 | name: habana-ai 9 | spec: 10 | image_registry: vault.habana.ai 11 | driver: 12 | driver_loader: 13 | images: 14 | ubuntu_22.04: 15 | repository: vault.habana.ai/habana-ai-operator/driver/ubuntu22.04/driver-installer 16 | tag: 1.19.1-26 17 | rhel_8.6: 18 | repository: vault.habana.ai/habana-ai-operator/driver/rhel8.6/driver-installer 19 | tag: 1.19.1-26 20 | rhel_9.2: 21 | repository: vault.habana.ai/habana-ai-operator/driver/rhel9.2/driver-installer 22 | tag: 1.19.1-26 23 | rhel_9.4: 24 | repository: vault.habana.ai/habana-ai-operator/driver/rhel9.4/driver-installer 25 | tag: 1.19.1-26 26 | tencentos_3.1: 27 | repository: vault.habana.ai/habana-ai-operator/driver/tencentos3.1/driver-installer 28 | tag: 1.19.1-26 29 | resources: 30 | limits: 31 | cpu: cpu_str_or_int_optional 32 | memory: memory_str_optional 33 | requests: 34 | cpu: cpu_str_or_int_optional 35 | memory: memory_str_optional 36 | repo_server: vault.habana.ai 37 | repo_path: artifactory/gaudi-installer/repos 38 | mlnx_ofed_repo_path: artifactory/gaudi-installer/deps 39 | mlnx_ofed_version: mlnx-ofed-5.8-2.0.3.0-rhel8.4-x86_64.tar.gz 40 | hugepages: hugepages_number_int_optional 41 | external_ports: turn_on_external_port_bool_optional 42 | firmware_flush: flush_firmware_on_the_gaudi_cards_bool_optional 43 | driver_runner: 44 | image: 45 | repository: vault.habana.ai/habana-ai-operator/driver/rhel9.4/driver-installer 46 | tag: 1.19.1-26 47 | resources: 48 | limits: 49 | cpu: cpu_str_or_int_optional 50 | memory: memory_str_optional 51 | requests: 52 | cpu: cpu_str_or_int_optional 53 | memory: memory_str_optional 54 | device_plugin: 55 | image: 56 | repository: vault.habana.ai/docker-k8s-device-plugin/docker-k8s-device-plugin 57 | tag: 1.19.1 58 | resources: 59 | limits: 60 | cpu: cpu_str_or_int_optional 61 | memory: memory_str_optional 62 | requests: 63 | cpu: cpu_str_or_int_optional 64 | memory: memory_str_optional 65 | runtime: 66 | runner: 67 | image: 68 | repository: vault.habana.ai/habana-ai-operator/habana-container-runtime 69 | tag: 1.19.1-26 70 | resources: 71 | limits: 72 | cpu: cpu_str_or_int_optional 73 | memory: memory_str_optional 74 | requests: 75 | cpu: cpu_str_or_int_optional 76 | memory: memory_str_optional 77 | configuration: 78 | container_engine: one_of_containerd_docker_crio 79 | engine_container_runtime_configuration: container_engine_configuration_optional 80 | habana_container_runtime_configuration: container_runtime_configuration_optional 81 | metric_exporter: 82 | runner: 83 | image: 84 | repository: vault.habana.ai/gaudi-metric-exporter/metric-exporter 85 | tag: 1.19.1-26 86 | resources: 87 | limits: 88 | cpu: cpu_str_or_int_optional 89 | memory: memory_str_optional 90 | requests: 91 | cpu: cpu_str_or_int_optional 92 | memory: memory_str_optional 93 | port: 41611 94 | interval: 20 95 | feature_discovery: 96 | runner: 97 | image: 98 | repository: vault.habana.ai/habana-ai-operator/habanalabs-feature-discovery 99 | tag: 1.19.1-26 100 | resources: 101 | limits: 102 | cpu: cpu_str_or_int_optional 103 | memory: memory_str_optional 104 | requests: 105 | cpu: cpu_str_or_int_optional 106 | memory: memory_str_optional 107 | nfd_plugin: boolean_nfd_installed 108 | bmc_monitoring: 109 | image: 110 | repository: vault.habana.ai/habana-bmc-exporter/bmc-exporter 111 | tag: 1.19.1-26 112 | resources: 113 | limits: 114 | cpu: cpu_str_or_int_optional 115 | memory: memory_str_optional 116 | requests: 117 | cpu: cpu_str_or_int_optional 118 | memory: memory_str_optional 119 | node_selector: 120 | key_optional: value_optional -------------------------------------------------------------------------------- /device_plugins/README.md: -------------------------------------------------------------------------------- 1 | # Setting up Intel Device Plugins Operator 2 | 3 | ## Overview 4 | Intel Device Plugins are utilized to advertise Intel hardware features (resources) to a Red Hat OpenShift Container Platform (RHOCP) Cluster. This allows workloads running on pods deployed within the clusters to leverage these features. To handle the deployment and lifecycle of these device plugins, the [Intel Device Plugins Operator](https://catalog.redhat.com/software/container-stacks/detail/61e9f2d7b9cdd99018fc5736) is used. The Intel Device Plugins container images and operator have been officially certified and published on the [Red Hat Ecosystem Catalog](https://catalog.redhat.com/software/container-stacks/detail/61e9f2d7b9cdd99018fc5736). For more details on the upstream project, please refer to [Intel Device Plugins for Kubernetes](https://github.com/intel/intel-device-plugins-for-kubernetes). 5 | 6 | ## Prerequisities 7 | - Provisioned RHOCP cluster. Follow steps [here](/README.md). 8 | - Setup Node Feature Discovery (NFD). Follow steps [here](/nfd/README.md). 9 | - Follow the additional prerequisites for provisioning Intel® Data Center GPU: 10 | - Setup out of tree drivers for Intel Data Center GPU provisioning. Follow the steps listed [here](/kmmo/README.md). 11 | - Follow the additional prerequisites for provisioning Intel® QuickAssist Technology: 12 | - Configure MCO for provisioning Intel QAT. Follow steps [here](/machine_configuration/README.md#machine-configuration-for-provisioning-intel-qat). 13 | 14 | ## Install Intel Device Plugins Operator on Red Hat OpenShift 15 | ### Installation via web console 16 | Follow the steps below to install Intel Device Plugins Operator using OpenShift web console: 17 | 1. In the OpenShift web console, navigate to **Operator** -> **OperatorHub**. 18 | 2. Search for **Intel Device Plugins Operator** in all items field -> Click **Install**. 19 | ### Verify Installation via web console 20 | 1. Go to **Operator** -> **Installed Operators**. 21 | 2. Verify that the status of the operator is **Succeeded**. 22 | 23 | ### Installation via command line interface (CLI) 24 | Apply the [install_operator.yaml](/device_plugins/install_operator.yaml) file: 25 | ``` 26 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/device_plugins/install_operator.yaml 27 | ``` 28 | 29 | ### Verify Installation via CLI 30 | Verify that the operator controller manager pod is up and running: 31 | ``` 32 | $ oc get pod | grep inteldeviceplugins-controller-manager 33 | 34 | inteldeviceplugins-controller-manager-6b8c76c867-hftqm 2/2 Running 0 17m 35 | ``` 36 | 37 | ## Resources Provided by Intel Device Plugins 38 | The resources are the user interface for customers to claim and consume the hardware features provided by Intel Device Plugins from the user pods. See below table for the details: 39 | 40 | | Feature | Resources | Description | Usage | 41 | | ------- | --------- | ----------- | ----- | 42 | | Intel® SGX | `sgx.intel.com/epc` | Intel SGX EPC memory for user pod to claim | [Link](https://github.com/intel/intel-technology-enabling-for-openshift/blob/64a6c86f3be25459c14ea988e892f9f5d873a8ca/tests/l2/sgx/sgx_job.yaml#L21) | 43 | | Intel® Data Center GPU Flex Series
Intel® Data Center GPU Max Series | `gpu.intel.com/i915 ` | Intel Data Center GPU Card for user pod to claim | [Link](https://github.com/intel/intel-technology-enabling-for-openshift/blob/main/device_plugins/deploy_gpu.md#using-intel-data-center-gpu-resource-exclusively) | 44 | | Intel® QAT | `qat.intel.com/cy`
`qat.intel.com/dc` | `cy`: Intel QAT VFIO Virtual Function device configured for cryptography for user pod to claim
`dc`: Intel QAT VFIO Virtual Function device configured for cryptography for user pod to claim | [Link](https://github.com/intel/intel-technology-enabling-for-openshift/blob/main/tests/l2/qat/qatlib_job.yaml#L24)
[Link](https://github.com/intel/intel-technology-enabling-for-openshift/blob/main/tests/l2/qat/qatlib_job.yaml#L28) | 45 | | Intel® DSA | `dsa.intel.com/wq-user-shared`
`dsa.intel.com/wq-user-dedicated` | Intel DSA Work Queue for user pod to claim | [Link](https://github.com/intel/intel-technology-enabling-for-openshift/blob/main/tests/l2/dsa/dsa_job.yaml#L27) | 46 | 47 | 48 | ## Creating Intel Device Plugin custom resource (CR) 49 | - To create an Intel SGX device plugin CR, follow this [link](/device_plugins/deploy_sgx.md). 50 | - To create an Intel GPU device plugin CR, follow this [link](/device_plugins/deploy_gpu.md). 51 | - To create an Intel QAT device plugin CR, follow this [link](/device_plugins/deploy_qat.md). 52 | - To create an Intel DSA device plugin CR, follow this [link](/device_plugins/deploy_dsa.md). 53 | -------------------------------------------------------------------------------- /.github/workflows/publish.yml: -------------------------------------------------------------------------------- 1 | name: Publish Documentation 2 | on: 3 | push: 4 | branches: 5 | - main 6 | - release-v1.3.0 7 | - release-v1.3.1 8 | - release-v1.4.0 9 | - release-v1.5.0 10 | - release-v1.5.1 11 | - release-v1.5.2 12 | - release-v1.6.0 13 | - release-v1.6.1 14 | 15 | permissions: 16 | contents: read 17 | 18 | jobs: 19 | build: 20 | permissions: 21 | contents: write # for Git to git push 22 | runs-on: ubuntu-22.04 23 | 24 | steps: 25 | - name: Install dependencies 26 | run: | 27 | sudo apt-get update 28 | sudo apt-get install -y python3-venv git 29 | - uses: actions/checkout@v4 30 | with: 31 | fetch-depth: 0 32 | ref: main 33 | - name: Set up doc directory 34 | run: | 35 | mkdir $HOME/output 36 | touch $HOME/output/.nojekyll 37 | echo "" >"$HOME/output/index.html" 38 | - name: Build devel 39 | run: | 40 | export GITHUB_SHA=$(git rev-parse HEAD) 41 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 42 | rm -rf _work/venv 43 | make vhtml 44 | mv _build/html $HOME/output/development 45 | - uses: actions/checkout@v4 46 | with: 47 | fetch-depth: 0 48 | ref: release-v1.3.0 49 | - name: Build release-v1.3.0 50 | run: | 51 | export GITHUB_SHA=$(git rev-parse HEAD) 52 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 53 | rm -rf _work/venv 54 | make vhtml 55 | mv _build/html $HOME/output/v1.3.0 56 | - uses: actions/checkout@v4 57 | with: 58 | fetch-depth: 0 59 | ref: release-v1.3.1 60 | - name: Build release-v1.3.1 61 | run: | 62 | export GITHUB_SHA=$(git rev-parse HEAD) 63 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 64 | rm -rf _work/venv 65 | make vhtml 66 | mv _build/html $HOME/output/v1.3.1 67 | - uses: actions/checkout@v4 68 | with: 69 | fetch-depth: 0 70 | ref: release-v1.4.0 71 | - name: Build release-v1.4.0 72 | run: | 73 | export GITHUB_SHA=$(git rev-parse HEAD) 74 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 75 | rm -rf _work/venv 76 | make vhtml 77 | mv _build/html $HOME/output/v1.4.0 78 | - uses: actions/checkout@v4 79 | with: 80 | fetch-depth: 0 81 | ref: release-v1.5.0 82 | - name: Build release-v1.5.0 83 | run: | 84 | export GITHUB_SHA=$(git rev-parse HEAD) 85 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 86 | rm -rf _work/venv 87 | make vhtml 88 | mv _build/html $HOME/output/v1.5.0 89 | - uses: actions/checkout@v4 90 | with: 91 | fetch-depth: 0 92 | ref: release-v1.5.1 93 | - name: Build release-v1.5.1 94 | run: | 95 | export GITHUB_SHA=$(git rev-parse HEAD) 96 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 97 | rm -rf _work/venv 98 | make vhtml 99 | mv _build/html $HOME/output/v1.5.1 100 | - uses: actions/checkout@v4 101 | with: 102 | fetch-depth: 0 103 | ref: release-v1.5.2 104 | - name: Build release-v1.5.2 105 | run: | 106 | export GITHUB_SHA=$(git rev-parse HEAD) 107 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 108 | rm -rf _work/venv 109 | make vhtml 110 | mv _build/html $HOME/output/v1.5.2 111 | - uses: actions/checkout@v4 112 | with: 113 | fetch-depth: 0 114 | ref: release-v1.6.0 115 | - name: Build release-v1.6.0 116 | run: | 117 | export GITHUB_SHA=$(git rev-parse HEAD) 118 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 119 | rm -rf _work/venv 120 | make vhtml 121 | mv _build/html $HOME/output/v1.6.0 122 | - uses: actions/checkout@v4 123 | with: 124 | fetch-depth: 0 125 | ref: release-v1.6.1 126 | - name: Build release-v1.6.1 127 | run: | 128 | export GITHUB_SHA=$(git rev-parse HEAD) 129 | export GITHUB_SHA_REF=$(git rev-parse --abbrev-ref HEAD) 130 | rm -rf _work/venv 131 | make vhtml 132 | mv _build/html $HOME/output/v1.6.1 133 | - name: Deploy the docs 134 | shell: bash 135 | env: 136 | GITHUB_TOKEN: ${{ secrets.GH_TOKEN }} 137 | run: | 138 | cd $HOME/output 139 | git init 140 | git config --global user.name "${GITHUB_ACTOR}" 141 | git config --global user.email "${GITHUB_ACTOR}@github.com" 142 | git add . 143 | git commit -m "latest html output" 144 | git push -f https://${GITHUB_ACTOR}:${GITHUB_TOKEN}@github.com/${GITHUB_REPOSITORY}.git HEAD:gh-pages -------------------------------------------------------------------------------- /tests/gaudi/l2/vllm_buildconfig.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | 4 | apiVersion: image.openshift.io/v1 5 | kind: ImageStream 6 | metadata: 7 | name: vllm-workload 8 | namespace: gaudi-validation 9 | spec: {} 10 | --- 11 | apiVersion: build.openshift.io/v1 12 | kind: BuildConfig 13 | metadata: 14 | name: vllm-workload 15 | namespace: gaudi-validation 16 | spec: 17 | triggers: 18 | - type: "ConfigChange" 19 | - type: "ImageChange" 20 | runPolicy: "Serial" 21 | source: 22 | type: Dockerfile 23 | git: 24 | uri: https://github.com/HabanaAI/vllm-fork.git 25 | ref: v1.19.1 26 | dockerfile: | 27 | ARG BASE_IMAGE=vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26 28 | FROM ${BASE_IMAGE} as habana-base 29 | 30 | USER root 31 | 32 | ENV VLLM_TARGET_DEVICE="hpu" 33 | ENV HABANA_SOFTWARE_VERSION="1.19.1" 34 | 35 | RUN dnf -y update --best --allowerasing --skip-broken && dnf clean all 36 | 37 | WORKDIR /workspace 38 | 39 | ## Python Installer ################################################################# 40 | FROM habana-base as python-install 41 | 42 | ARG PYTHON_VERSION=3.11 43 | 44 | ENV VIRTUAL_ENV=/opt/vllm 45 | ENV PATH="$VIRTUAL_ENV/bin:$PATH" 46 | RUN dnf install -y --setopt=install_weak_deps=0 --nodocs \ 47 | python${PYTHON_VERSION}-wheel && \ 48 | python${PYTHON_VERSION} -m venv $VIRTUAL_ENV --system-site-packages && pip install --no-cache -U pip wheel && dnf clean all 49 | 50 | ## Python Habana base ################################################################# 51 | FROM python-install as python-habana-base 52 | 53 | ENV VIRTUAL_ENV=/opt/vllm 54 | ENV PATH="$VIRTUAL_ENV/bin:$PATH" 55 | 56 | # install Habana Software and common dependencies 57 | RUN --mount=type=cache,target=/root/.cache/pip \ 58 | --mount=type=bind,source=requirements-common.txt,target=requirements-common.txt \ 59 | --mount=type=bind,source=requirements-hpu.txt,target=requirements-hpu.txt \ 60 | pip install \ 61 | -r requirements-hpu.txt 62 | 63 | ## Builder ##################################################################### 64 | FROM python-habana-base AS build 65 | 66 | # install build dependencies 67 | 68 | # copy input files 69 | COPY csrc csrc 70 | COPY setup.py setup.py 71 | COPY cmake cmake 72 | COPY CMakeLists.txt CMakeLists.txt 73 | COPY requirements-common.txt requirements-common.txt 74 | COPY requirements-hpu.txt requirements-hpu.txt 75 | COPY pyproject.toml pyproject.toml 76 | 77 | # max jobs used by Ninja to build extensions 78 | ARG max_jobs=2 79 | ENV MAX_JOBS=${max_jobs} 80 | # # make sure punica kernels are built (for LoRA) 81 | # HPU currently doesn't support LoRA 82 | # ENV VLLM_INSTALL_PUNICA_KERNELS=1 83 | 84 | # Copy the entire directory before building wheel 85 | COPY vllm vllm 86 | 87 | ENV CCACHE_DIR=/root/.cache/ccache 88 | RUN --mount=type=cache,target=/root/.cache/ccache \ 89 | --mount=type=cache,target=/root/.cache/pip \ 90 | --mount=type=bind,src=.git,target=/workspace/.git \ 91 | env CFLAGS="-march=haswell" \ 92 | CXXFLAGS="$CFLAGS $CXXFLAGS" \ 93 | CMAKE_BUILD_TYPE=Release \ 94 | python3 setup.py bdist_wheel --dist-dir=dist 95 | 96 | ## Release ##################################################################### 97 | FROM python-install AS vllm-openai 98 | 99 | WORKDIR /workspace 100 | 101 | ENV VIRTUAL_ENV=/opt/vllm 102 | ENV PATH=$VIRTUAL_ENV/bin/:$PATH 103 | 104 | # Triton needs a CC compiler 105 | RUN dnf install -y --setopt=install_weak_deps=0 --nodocs gcc \ 106 | && dnf clean all 107 | 108 | # install vllm wheel first, so that torch etc will be installed 109 | RUN --mount=type=bind,from=build,src=/workspace/dist,target=/workspace/dist \ 110 | --mount=type=cache,target=/root/.cache/pip \ 111 | pip install $(echo dist/*.whl)'[tensorizer]' --verbose 112 | 113 | ENV HF_HUB_OFFLINE=1 \ 114 | PORT=8000 \ 115 | HOME=/home/vllm \ 116 | VLLM_USAGE_SOURCE=production-docker-image 117 | 118 | # setup non-root user for OpenShift 119 | # In OpenShift the user ID is randomly assigned, for compatibility we also 120 | # set up a non-root user here. 121 | RUN umask 002 \ 122 | && useradd --uid 2000 --gid 0 vllm \ 123 | && chmod g+rwx $HOME /usr/src /workspace 124 | 125 | COPY LICENSE /licenses/vllm.md 126 | 127 | USER 2000 128 | ENTRYPOINT ["python3", "-m", "vllm.entrypoints.openai.api_server"] 129 | strategy: 130 | type: Docker 131 | noCache: true 132 | dockerStrategy: 133 | buildArgs: 134 | - name: "BASE_IMAGE" 135 | value: "vault.habana.ai/gaudi-docker/1.19.1/rhel9.4/habanalabs/pytorch-installer-2.5.1:1.19.1-26" 136 | output: 137 | to: 138 | kind: ImageStreamTag 139 | name: vllm-workload:latest -------------------------------------------------------------------------------- /conf.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | # 3 | # Configuration file for the Sphinx documentation builder. 4 | # 5 | # This file does only contain a selection of the most common options. For a 6 | # full list see the documentation: 7 | # http://www.sphinx-doc.org/en/master/config 8 | 9 | # -- Path setup -------------------------------------------------------------- 10 | 11 | # If extensions (or modules to document with autodoc) are in another directory, 12 | # add these directories to sys.path here. If the directory is relative to the 13 | # documentation root, use os.path.abspath to make it absolute, like shown here. 14 | # 15 | # import os 16 | # import sys 17 | # sys.path.insert(0, os.path.abspath('.')) 18 | 19 | # -- Project information ----------------------------------------------------- 20 | 21 | project = 'Intel® Technology Enabling for OpenShift*' 22 | copyright = '2024, Intel® Corporation' 23 | author = 'Intel® Corporation' 24 | 25 | # The short X.Y version 26 | # version = 'devel' 27 | # The full version, including alpha/beta/rc tags 28 | # release = 'GA' 29 | 30 | 31 | # --------------------------------- 32 | # Reference for sphinx_md : https://pypi.org/project/sphinx-md/ 33 | # --------------------------------- 34 | from os import getenv 35 | 36 | baseBranch = "main" 37 | sphinx_md_useGitHubURL = True 38 | commitSHA = getenv('GITHUB_SHA') 39 | githubBaseURL = 'https://github.com/' + (getenv('GITHUB_REPOSITORY') or 'intel/intel-technology-enabling-for-openshift') + '/' 40 | githubFileURL = githubBaseURL + "blob/" 41 | githubDirURL = githubBaseURL + "tree/" 42 | if commitSHA: 43 | githubFileURL = githubFileURL + commitSHA + "/" 44 | githubDirURL = githubDirURL + commitSHA + "/" 45 | else: 46 | githubFileURL = githubFileURL + baseBranch + "/" 47 | githubDirURL = githubDirURL + baseBranch + "/" 48 | sphinx_md_githubFileURL = githubFileURL 49 | sphinx_md_githubDirURL = githubDirURL 50 | 51 | # Version displayed in the upper left corner 52 | # This value is set in the github workflow environment 53 | commitREF = getenv('GITHUB_SHA_REF', default = "unknown") 54 | if commitREF.startswith("release-"): 55 | version = commitREF[len("release-"):].strip() 56 | else: 57 | version = "development" 58 | 59 | 60 | # Versions list with URLs using tags displayed in the lower left corner 61 | from git import Repo 62 | versions_to_exclude = set(['v1.0.0', 'v1.0.1','v1.1.0', 'v1.2.0', 'v1.2.1']) 63 | repo = Repo( search_parent_directories=True ) 64 | github_repo = "/intel-technology-enabling-for-openshift/" 65 | release_versions = [("development", github_repo)] 66 | tags = reversed([tag.name for tag in repo.tags]) 67 | release_versions.extend((str(tag), github_repo + tag) for tag in tags if str(tag) not in versions_to_exclude) 68 | 69 | # -- General configuration --------------------------------------------------- 70 | 71 | # If your documentation needs a minimal Sphinx version, state it here. 72 | # 73 | # needs_sphinx = '1.0' 74 | 75 | # Add any Sphinx extension module names here, as strings. They can be 76 | # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom 77 | # ones. 78 | extensions = ['myst_parser', 'sphinx_md', ] 79 | myst_heading_anchors = 5 80 | # myst_enable_extensions = [ 81 | # "html_admonition", 82 | # ] 83 | # Add any paths that contain templates here, relative to this directory. 84 | templates_path = ['_templates'] 85 | 86 | # The suffix(es) of source filenames. 87 | # You can specify multiple suffix as a list of string: 88 | # 89 | source_suffix = ['.rst', '.md'] 90 | 91 | # List of patterns, relative to source directory, that match files and 92 | # directories to ignore when looking for source files. 93 | # This pattern also affects html_static_path and html_extra_path. 94 | exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store'] 95 | 96 | # -- Options for HTML output ------------------------------------------------- 97 | 98 | # The theme to use for HTML and HTML Help pages. See the documentation for 99 | # a list of builtin themes. 100 | # 101 | html_theme = 'sphinx_rtd_theme' 102 | html_title = "Intel® Technology Enabling for OpenShift*" 103 | # Theme options are theme-specific and customize the look and feel of a theme 104 | # further. For a list of options available for each theme, see the 105 | # documentation. 106 | # 107 | html_theme_options = { 108 | "display_version": True, 109 | } 110 | 111 | html_context = { 112 | 'display_github': True, 113 | 'github_host': 'github.com', 114 | 'github_user': 'intel', 115 | 'github_repo': 'intel-technology-enabling-for-openshift', 116 | 'github_version': 'main/', 117 | 'versions_menu': True, 118 | 'version': version, 119 | 'versions': release_versions, 120 | } 121 | html_css_files = [ 122 | 'custom.css', 123 | ] 124 | 125 | # Add any paths that contain custom static files (such as style sheets) here, 126 | # relative to this directory. They are copied after the builtin static files, 127 | # so a file named "default.css" will overwrite the builtin "default.css". 128 | 129 | html_static_path = ['_static'] 130 | 131 | 132 | # Custom sidebar templates, must be a dictionary that maps document names 133 | # to template names. 134 | # 135 | # The default sidebars (for documents that don't match any pattern) are 136 | # defined by theme itself. Builtin themes are using these templates by 137 | # default: ``['localtoc.html', 'relations.html', 'sourcelink.html', 138 | # 'searchbox.html']``. 139 | # 140 | # html_sidebars = {} 141 | 142 | 143 | # -- Options for HTMLHelp output --------------------------------------------- 144 | 145 | # Output file base name for HTML help builder. 146 | htmlhelp_basename = 'IntelTechnologyEnablingforOpenShiftdoc' -------------------------------------------------------------------------------- /CODE_OF_CONDUCT.md: -------------------------------------------------------------------------------- 1 | # Contributor Covenant Code of Conduct 2 | 3 | ## Our Pledge 4 | 5 | We as members, contributors, and leaders pledge to make participation in our 6 | community a harassment-free experience for everyone, regardless of age, body 7 | size, visible or invisible disability, ethnicity, sex characteristics, gender 8 | identity and expression, level of experience, education, socio-economic status, 9 | nationality, personal appearance, race, caste, color, religion, or sexual 10 | identity and orientation. 11 | 12 | We pledge to act and interact in ways that contribute to an open, welcoming, 13 | diverse, inclusive, and healthy community. 14 | 15 | ## Our Standards 16 | 17 | Examples of behavior that contributes to a positive environment for our 18 | community include: 19 | 20 | * Demonstrating empathy and kindness toward other people 21 | * Being respectful of differing opinions, viewpoints, and experiences 22 | * Giving and gracefully accepting constructive feedback 23 | * Accepting responsibility and apologizing to those affected by our mistakes, 24 | and learning from the experience 25 | * Focusing on what is best not just for us as individuals, but for the overall 26 | community 27 | 28 | Examples of unacceptable behavior include: 29 | 30 | * The use of sexualized language or imagery, and sexual attention or advances of 31 | any kind 32 | * Trolling, insulting or derogatory comments, and personal or political attacks 33 | * Public or private harassment 34 | * Publishing others' private information, such as a physical or email address, 35 | without their explicit permission 36 | * Other conduct which could reasonably be considered inappropriate in a 37 | professional setting 38 | 39 | ## Enforcement Responsibilities 40 | 41 | Community leaders are responsible for clarifying and enforcing our standards of 42 | acceptable behavior and will take appropriate and fair corrective action in 43 | response to any behavior that they deem inappropriate, threatening, offensive, 44 | or harmful. 45 | 46 | Community leaders have the right and responsibility to remove, edit, or reject 47 | comments, commits, code, wiki edits, issues, and other contributions that are 48 | not aligned to this Code of Conduct, and will communicate reasons for moderation 49 | decisions when appropriate. 50 | 51 | ## Scope 52 | 53 | This Code of Conduct applies within all community spaces, and also applies when 54 | an individual is officially representing the community in public spaces. 55 | Examples of representing our community include using an official e-mail address, 56 | posting via an official social media account, or acting as an appointed 57 | representative at an online or offline event. 58 | 59 | ## Enforcement 60 | 61 | Instances of abusive, harassing, or otherwise unacceptable behavior may be 62 | reported to the community leaders responsible for enforcement at 63 | CommunityCodeOfConduct AT intel DOT com. 64 | All complaints will be reviewed and investigated promptly and fairly. 65 | 66 | All community leaders are obligated to respect the privacy and security of the 67 | reporter of any incident. 68 | 69 | ## Enforcement Guidelines 70 | 71 | Community leaders will follow these Community Impact Guidelines in determining 72 | the consequences for any action they deem in violation of this Code of Conduct: 73 | 74 | ### 1. Correction 75 | 76 | **Community Impact**: Use of inappropriate language or other behavior deemed 77 | unprofessional or unwelcome in the community. 78 | 79 | **Consequence**: A private, written warning from community leaders, providing 80 | clarity around the nature of the violation and an explanation of why the 81 | behavior was inappropriate. A public apology may be requested. 82 | 83 | ### 2. Warning 84 | 85 | **Community Impact**: A violation through a single incident or series of 86 | actions. 87 | 88 | **Consequence**: A warning with consequences for continued behavior. No 89 | interaction with the people involved, including unsolicited interaction with 90 | those enforcing the Code of Conduct, for a specified period of time. This 91 | includes avoiding interactions in community spaces as well as external channels 92 | like social media. Violating these terms may lead to a temporary or permanent 93 | ban. 94 | 95 | ### 3. Temporary Ban 96 | 97 | **Community Impact**: A serious violation of community standards, including 98 | sustained inappropriate behavior. 99 | 100 | **Consequence**: A temporary ban from any sort of interaction or public 101 | communication with the community for a specified period of time. No public or 102 | private interaction with the people involved, including unsolicited interaction 103 | with those enforcing the Code of Conduct, is allowed during this period. 104 | Violating these terms may lead to a permanent ban. 105 | 106 | ### 4. Permanent Ban 107 | 108 | **Community Impact**: Demonstrating a pattern of violation of community 109 | standards, including sustained inappropriate behavior, harassment of an 110 | individual, or aggression toward or disparagement of classes of individuals. 111 | 112 | **Consequence**: A permanent ban from any sort of public interaction within the 113 | community. 114 | 115 | ## Attribution 116 | 117 | This Code of Conduct is adapted from the [Contributor Covenant][homepage], 118 | version 2.1, available at 119 | [https://www.contributor-covenant.org/version/2/1/code_of_conduct.html][v2.1]. 120 | 121 | Community Impact Guidelines were inspired by 122 | [Mozilla's code of conduct enforcement ladder][Mozilla CoC]. 123 | 124 | For answers to common questions about this code of conduct, see the FAQ at 125 | [https://www.contributor-covenant.org/faq][FAQ]. Translations are available at 126 | [https://www.contributor-covenant.org/translations][translations]. 127 | 128 | [homepage]: https://www.contributor-covenant.org 129 | [v2.1]: https://www.contributor-covenant.org/version/2/1/code_of_conduct.html 130 | [Mozilla CoC]: https://github.com/mozilla/diversity 131 | [FAQ]: https://www.contributor-covenant.org/faq 132 | -------------------------------------------------------------------------------- /gaudi/README.md: -------------------------------------------------------------------------------- 1 | # Setting up Intel Gaudi AI Accelerator Operator 2 | 3 | ## Overview 4 | [Intel Gaudi AI Accelerator Operator](https://catalog.redhat.com/software/container-stacks/detail/6683b2cce45daa25e36bddcb) is used to provision Intel Gaudi Accelerator with OpenShift. The steps and yaml files mentioned in this document to provision the Gaudi accelerator are based on [Intel Gaudi AI Accelerator Operator for OpenShift](https://docs.habana.ai/en/latest/Orchestration/Intel_Gaudi_Base_Operator/index.html). 5 | 6 | If you are familiar with the steps here to manually provision the accelerator, the Red Hat certified Operator and Ansible based [One-Click](/one_click/README.md#reference-playbook-–-habana-gaudi-provisioning) solution can be used as a reference to provision the accelerator automatically. 7 | 8 | ## Prerequisities 9 | - To Provision RHOCP cluster, follow steps [here](/README.md#provisioning-rhocp-cluster). 10 | 11 | ## Install Intel Gaudi AI Accelerator Operator on Red Hat OpenShift 12 | ### Installation via web console 13 | Follow the steps below to install Intel Gaudi AI Accelerator Operator using OpenShift web console: 14 | 1. In the OpenShift web console, navigate to **Operator** -> **OperatorHub**. 15 | 2. Search for **Intel Gaudi AI Accelerator Operator** in all items field -> Click **Install**. 16 | ### Verify Installation via web console 17 | 1. Go to **Operator** -> **Installed Operators**. 18 | 2. Verify that the status of the operator is **Succeeded**. 19 | 20 | ### Installation via Command Line Interface (CLI) 21 | ``` 22 | oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/gaudi/gaudi_install_operator.yaml 23 | ``` 24 | 25 | ### Verify Installation via CLI 26 | Verify that the operator controller manager pod is up and running: 27 | ``` 28 | oc get pods -n habana-ai-operator 29 | 30 | NAME READY STATUS RESTARTS AGE 31 | controller-manager-6c8459d9cb-fqs8h 2/2 Running 0 25m 32 | ``` 33 | 34 | ## Creating Intel Gaudi AI Accelerator Operator ClusterPolicy Instance 35 | To create a Habana Gaudi device plugin CR, follow the steps below. 36 | 37 | ### Create CR via web console 38 | 1. Go to **Operator** -> **Installed Operators**. 39 | 2. Open **Intel Gaudi AI Accelerator Operator**. 40 | 3. Navigate to tab **Cluster Policy**. 41 | 4. Click **Create ClusterPolicy** -> set correct parameters -> Click **Create**. To set correct parameters please refer [Using RedHat OpenShift Container Platform Console](https://docs.habana.ai/en/latest/Installation_Guide/Additional_Installation/Kubernetes_Installation/Kubernetes_Operator.html#id1). 42 | 43 | ### Verify via web console 44 | 1. Verify CR by checking the status of **Workloads** -> **DaemonSet** -> **habana-ai-device-plugin-ds**, **habana-ai-driver-rhel-9-4-xxxxx**, **habana-ai-feature-discovery-ds**, **habana-ai-metric-exporter-ds**, **habana-ai-runtime-ds**. 45 | 2. Now `ClusterPolicy` is created. 46 | 47 | ### Create CR via CLI 48 | Apply the CR yaml file: 49 | ``` 50 | oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/gaudi/gaudi_cluster_policy.yaml 51 | ``` 52 | 53 | ### Verify the ClusterPolicy CR is created 54 | You can use command below to verify that the `ClusterPolicy` CR has been created: 55 | ``` 56 | oc get pod -n habana-ai-operator 57 | 58 | NAME READY STATUS RESTARTS AGE 59 | habana-ai-device-plugin-ds-thj7b 1/1 Running 0 10d 60 | habana-ai-driver-rhel-9-4-416-94-202412170927-0-ds-vqhzb 1/1 Running 2 10d 61 | habana-ai-feature-discovery-ds-ztl2j 1/1 Running 5 10d 62 | habana-ai-metric-exporter-ds-g5qqh 1/1 Running 0 10d 63 | habana-ai-operator-controller-manager-6c995b5646-wl7cp 2/2 Running 0 10d 64 | habana-ai-runtime-ds-x49lf 1/1 Running 0 10d 65 | ``` 66 | Alternatively, you can also check the status of the `ClusterPolicy` CR like below: 67 | ``` 68 | oc describe ClusterPolicy habana-ai -n habana-ai-operator 69 | 70 | Name: habana-ai 71 | Namespace: habana-ai-operator 72 | . 73 | . 74 | Status: 75 | Conditions: 76 | Last Transition Time: 2025-01-21T18:50:46Z 77 | Message: All resources have been successfully reconciled 78 | Reason: Reconciled 79 | Status: True 80 | ``` 81 | ## Verify Gaudi Provisioning 82 | After the `ClusterPolicy` instance CR is created, it will take some time for the operator to download the Gaudi OOT driver source code and build it on-premise with the help of the KMM operator. The OOT driver module binaries will be loaded into the RHCOS kernel on each node with Gaudi cards labelled by feature discovery. Then, the Gaudi device plugin can advertise the Gaudi resources listed in the table for the pods on OpenShit to use. Run the command below to check the availability of Gaudi resources: 83 | ``` 84 | oc describe node | grep habana.ai/gaudi 85 | 86 | habana.ai/gaudi: 8 -> Gaudi cards number on the cluster 87 | habana.ai/gaudi: 8 -> Gaudi cards number allocatble on the cluster 88 | habana.ai/gaudi 4 4 -> number of Gaudi cards allocated and number of Gardi cards available 89 | ``` 90 | 91 | To view the metrics on a node with Gaudi card, refer [Collecting Metrics](https://docs.habana.ai/en/latest/Orchestration/Prometheus_Metric_Exporter.html?highlight=metrics#collecting-metrics). 92 | 93 | ## Resources Provided by Habana Gaudi Device Plugin 94 | The resources provided are the user interface for customers to claim and consume the hardware features from the user pods. See below table for the details: 95 | 96 | | Feature | Resources | Description | 97 | | ------- | --------- | ----------- | 98 | | Habana Gaudi | `habana.ai/gaudi` | Number of Habana Gaudi Card resources ready to claim | 99 | 100 | ## Upgrade Intel Gaudi SPI Firmware 101 | Refer [Upgrade Intel Gaudi SPI Firmware](/gaudi/Gaudi-SPI-Firmware-Upgrade.md) to upgrade the SPI Firmware on Intel Gaudi. -------------------------------------------------------------------------------- /workloads/opea/chatqna/README.md: -------------------------------------------------------------------------------- 1 | # Deploy OPEA ChatQnA workload on OCP 2 | 3 | ## Overview 4 | The workload is based on the [OPEA ChatQnA Application](https://github.com/opea-project/GenAIExamples/tree/v0.8/ChatQnA) running on Intel® Gaudi Accelerator with OpenShift and OpenShift AI. Refer to the [OPEA Generative AI Examples](https://github.com/opea-project/GenAIExamples/tree/v0.8) for more details about the OPEA workloads. 5 | 6 | **Note**: It is still under heavy development, and the updates are expected. 7 | 8 | ## Prerequisites 9 | * Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster) 10 | * The Persistent storage using NFS is ready. Refer to [documentation](https://docs.openshift.com/container-platform/4.16/storage/persistent_storage/persistent-storage-nfs.html) for the details to set it up. 11 | 12 | **Note**: Refer to [documentation](https://docs.openshift.com/container-platform/4.16/storage/index.html) for setting up other types of persistent storages. 13 | * Provisioned Intel Gaudi accelerator on RHOCP cluster. Follow steps [here](/gaudi/README.md) 14 | * RHOAI is installed. Follow steps [here](/e2e/inference/README.md/#install-rhoai) 15 | * The Intel Gaudi AI accelerator is enabled with RHOAI. Follow steps [here](/e2e/inference/README.md/#enable-intel-gaudi-ai-accelerator-with-rhoai) 16 | * Minio based S3 service ready for RHOAI. Follow steps [here](https://ai-on-openshift.io/tools-and-applications/minio/minio/#create-a-matching-data-connection-for-minio) 17 | 18 | ## Deploy Model Serving for OPEA ChatQnA Microservices with RHOAI 19 | 20 | ### Create OpenShift AI Data Science Project 21 | 22 | * Click ```Search -> Routes -> rhods-dashboard``` from the OCP web console and launch the RHOAI dashboard. 23 | 24 | * Follow the dashboard and click ```Data Science Projects``` to create a project. For example, ```OPEA-chatqna-modserving```. 25 | 26 | ### Preload the models 27 | 28 | * Refer to [link](https://huggingface.co/docs/hub/en/models-downloading) and download the model [Llama2-70b-chat-hf](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf). 29 | 30 | * Refer to [link](https://ai-on-openshift.io/tools-and-applications/minio/minio/#create-a-matching-data-connection-for-minio) and upload the model to minio/s3 storage. 31 | 32 | * Click ```OPEA-chatqna-modserving```, and choose ```Data Connection``` section. In the fields, add your access and secret keys from minio. Follow [link](https://ai-on-openshift.io/tools-and-applications/minio/minio/#create-a-matching-data-connection-for-minio). 33 | 34 | ### Launch the Model Serving with Intel Gaudi AI Accelerator 35 | 36 | * Click on the Settings and choose ```ServingRuntime```. Copy or import the [tgi_gaudi_servingruntime.yaml](tgi_gaudi_servingruntime.yaml). The [tgi-gaudi](https://github.com/huggingface/tgi-gaudi) serving runtime is used. Follow the image below. 37 | 38 | ![Alt text](/docs/images/tgi-serving-runtime.png) 39 | 40 | * In the project ```OPEA-chatqna-modserving``` --> ```Models``` section and follow the image below. 41 | 42 | ![Alt text](/docs/images/rhoai-deploy-model.png) 43 | 44 | * The model server is now in the creation state. Once ready, the status will be updated to green and the inference endpoint can be seen. Refer to the image below. 45 | 46 | ![Alt text](/docs/images/model-server-status.png) 47 | 48 | ## Deploy ChatQnA Megaservice and Database 49 | 50 | ### Create namespace 51 | 52 | ``` 53 | oc create namespace opea-chatqna 54 | ``` 55 | 56 | ### Create persistent volumes 57 | The NFS is used to create the Persistent Volumes for ChatQnA MegaService to claim and use. 58 | 59 | Make sure to update NFS server IP and path in ```persistent_volumes.yaml``` before applying command below. 60 | For example: 61 | ``` 62 | nfs: 63 | server: 10.20.1.2 # nfs server 64 | path: /my_nfs # nfs path 65 | ``` 66 | 67 | ``` 68 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/workloads/opea/chatqna/persistent_volumes.yaml 69 | 70 | ``` 71 | 72 | * Check that the persistent volumes are created: 73 | 74 | ``` 75 | $ oc get pv 76 | NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS 77 | chatqna-megaservice-pv-0 100Mi RWO Retain Available 78 | chatqna-megaservice-pv-1 100Mi RWO Retain Available 79 | chatqna-megaservice-pv-2 100Mi RWO Retain Available 80 | 81 | ``` 82 | ### Building OPEA ChatQnA MegaService Container Image 83 | ``` 84 | create_megaservice_container.sh 85 | ``` 86 | 87 | ### Deploy Redis Vector Database Service 88 | ``` 89 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/workloads/opea/chatqna/redis_deployment_service.yaml 90 | 91 | ``` 92 | 93 | Check that the pod and service are running: 94 | 95 | ``` 96 | $ oc get pods 97 | NAME READY STATUS RESTARTS AGE 98 | redis-vector-db-6b5747bf7-sl8fr 1/1 Running 0 21s 99 | ``` 100 | 101 | ``` 102 | $ oc get svc 103 | NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE 104 | redis-vector-db ClusterIP 1.2.3.4 6379/TCP,8001/TCP 43s 105 | ``` 106 | 107 | ### Deploy ChatQnA MegaService 108 | 109 | Update the inference endpoint from the in the chatqna_megaservice_deployment. 110 | 111 | ``` 112 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/workloads/opea/chatqna/chatqna_megaservice_deployment.yaml 113 | ``` 114 | 115 | Check that the pod and service are running: 116 | 117 | ``` 118 | $ oc get pods 119 | NAME READY STATUS RESTARTS AGE 120 | chatqna-megaservice-54487649b5-sgsh2 1/1 Running 0 95s 121 | ``` 122 | 123 | ``` 124 | $ oc get svc 125 | NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE 126 | chatqna-megaservice ClusterIP 1.2.3.4 8000/TCP 99s 127 | ``` 128 | 129 | ### Verify the Megaservice 130 | Use the command below: 131 | 132 | ``` 133 | curl /v1/rag/chat_stream \ 134 | -X POST \ 135 | -d '{"query":"What is a constellation?"}' \ 136 | -H 'Content-Type: application/json' 137 | 138 | ``` -------------------------------------------------------------------------------- /kmmo/README.md: -------------------------------------------------------------------------------- 1 | # Setting up Out of Tree Drivers 2 | 3 | ## Introduction 4 | [Kernel module management (KMM) operator](https://github.com/rh-ecosystem-edge/kernel-module-management) manages the deployment and lifecycle of out-of-tree kernel modules on RHOCP. 5 | 6 | In this release, KMM operator is used to manage and deploy the Intel® Data Center GPU driver container image on the RHOCP cluster. 7 | 8 | Intel data center GPU driver container images are released from [Intel Data Center GPU Driver for OpenShift Project](https://github.com/intel/intel-data-center-gpu-driver-for-openshift/tree/main/release#intel-data-center-gpu-driver-container-images-for-openshift-release). 9 | 10 | ## KMM operator working mode 11 | - **Pre-build mode** - This is the default and recommended mode. KMM Operator uses [this pre-built and certified Intel Data Center GPU driver container image](https://catalog.redhat.com/software/containers/intel/intel-data-center-gpu-driver-container/6495ee55c8b2461e35fb8264), which is published on the Red Hat Ecosystem Catalog to provision Intel Data Center GPUs on a RHOCP cluster. 12 | - **On-premises build mode** - Users can optionally build and deploy their own driver container images on-premises through the KMM operator. 13 | 14 | ## Prerequisites 15 | - Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster). 16 | - Setup node feature discovery. Follow steps [here](/nfd/README.md). 17 | 18 | ## Install KMM operator 19 | Follow the installation guide below to install the KMM operator via CLI or web console. 20 | - [Install from CLI](https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/specialized_hardware_and_driver_enablement/kernel-module-management-operator#kmm-install-using-cli_kernel-module-management-operator) 21 | - [Install from web console](https://docs.redhat.com/en/documentation/openshift_container_platform/4.18/html/specialized_hardware_and_driver_enablement/kernel-module-management-operator#kmm-install-using-web-console_kernel-module-management-operator) 22 | 23 | ## Canary deployment with KMM 24 | Canary deployment is enabled by default to deploy the driver container image only on specific node(s) to ensure the initial deployment succeeds prior to rollout to all the eligible nodes in the cluster. This safety mechanism can reduce risk and prevent a deployment from adversely affecting the entire cluster. 25 | 26 | ## Set alternative firmware path at runtime with KMM 27 | **NOTE**: This update is required only when using KMM version of v2.1.1 or lower. Starting with v2.2.0, it is not required. 28 | 29 | Follow the steps below to set the alternative firmware path at runtime. 30 | 31 | 1. Update KMM operator `ConfigMap` to set `worker.firmwareHostPath` to `/var/lib/firmware` 32 | 33 | ``` 34 | $ oc patch configmap kmm-operator-manager-config -n openshift-kmm --type='json' -p='[{"op": "add", "path": "/data/controller_config.yaml", "value": "healthProbeBindAddress: :8081\nmetricsBindAddress: 127.0.0.1:8080\nleaderElection:\n enabled: true\n resourceID: kmm.sigs.x-k8s.io\nwebhook:\n disableHTTP2: true\n port: 9443\nworker:\n runAsUser: 0\n seLinuxType: spc_t\n firmwareHostPath: /var/lib/firmware"}]' 35 | ``` 36 | 37 | 2. Delete the KMM operator controller pod for `ConfigMap` changes to take effect. 38 | ``` 39 | $ oc get pods -n openshift-kmm | grep -i "kmm-operator-controller-" | awk '{print $1}' | xargs oc delete pod -n openshift-kmm 40 | ``` 41 | 42 | For more details, see [link.](https://openshift-kmm.netlify.app/documentation/firmwares/#setting-the-kernels-firmware-search-path) 43 | 44 | ## Deploy Intel Data Center GPU Driver with pre-build mode 45 | Follow the steps below to deploy the driver container image with pre-build mode. 46 | 1. Find all nodes with an Intel Data Center GPU card using the following command: 47 | ``` 48 | $ oc get nodes -l intel.feature.node.kubernetes.io/gpu=true 49 | ``` 50 | Example output: 51 | ``` 52 | NAME STATUS ROLES AGE VERSION 53 | icx-dgpu-1 Ready worker 30d v1.25.4+18eadca 54 | ``` 55 | 56 | 2. Label the node(s) in the cluster using the command shown below for the initial canary deployment. 57 | ``` 58 | $ oc label node intel.feature.node.kubernetes.io/dgpu-canary=true 59 | ``` 60 | 61 | 3. Use pre-build mode to deploy the driver container. 62 | ``` 63 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/kmmo/intel-dgpu.yaml 64 | ``` 65 | 66 | 4. After the driver is verified on the cluster through the canary deployment, simply remove the line shown below from the [`intel-dgpu.yaml`](/kmmo/intel-dgpu.yaml) file and reapply the yaml file to deploy the driver to the entire cluster. As a cluster administrator, you can also select another deployment policy. 67 | ``` 68 | intel.feature.node.kubernetes.io/dgpu-canary: 'true' 69 | ``` 70 | 71 | ## Verification 72 | To verify that the drivers have been loaded, follow the steps below: 73 | 1. List the nodes labeled with `kmm.node.kubernetes.io/openshift-kmm.intel-dgpu.ready` using the command shown below: 74 | ``` 75 | $ oc get nodes -l kmm.node.kubernetes.io/openshift-kmm.intel-dgpu.ready 76 | ``` 77 | Example output: 78 | ``` 79 | NAME STATUS ROLES AGE VERSION 80 | icx-dgpu-1 Ready worker 30d v1.25.4+18eadca 81 | ``` 82 | The label shown above indicates that the KMM operator has successfully deployed the drivers and firmware on the node. 83 | 84 | 2. If you want to further debug the driver on the node, follow these steps: 85 | a. Navigate to the web console (Compute -> Nodes -> Select a node that has the GPU card -> Terminal). 86 | b. Run the commands shown below in the web console terminal: 87 | ``` 88 | $ chroot /host 89 | $ lsmod | grep i915 90 | ``` 91 | Ensure `i915` and `intel_vsec` are loaded in the kernel, as shown in the output below: 92 | ``` 93 | i915 3633152 0 94 | i915_compat 16384 1 i915 95 | intel_vsec 16384 1 i915 96 | intel_gtt 20480 1 i915 97 | video 49152 1 i915 98 | i2c_algo_bit 16384 1 i915 99 | drm_kms_helper 290816 1 i915 100 | drm 589824 3 drm_kms_helper,i915 101 | dmabuf 77824 4 drm_kms_helper,i915,i915_compat,dr 102 | ``` 103 | c. Run dmesg to ensure there are no errors in the kernel message log. 104 | 105 | ## See Also 106 | -------------------------------------------------------------------------------- /e2e/inference/README.md: -------------------------------------------------------------------------------- 1 | # Intel AI Inference End-to-End Solution 2 | 3 | ## Overview 4 | Intel AI inference end-to-end solution with RHOCP is based on the Intel® Data Center GPU Flex Series provisioning, Intel® OpenVINO™, and [Red Hat OpenShift AI](https://www.redhat.com/en/technologies/cloud-computing/openshift/openshift-ai) (RHOAI) on RHOCP. There are two AI inference modes verified with Intel® Xeon® processors and Intel Data Center GPU Flex Series with RHOCP. 5 | * Interactive mode – RHOAI provides OpenVINO based Jupyter Notebooks for users to interactively debug the inference applications or [optimize the models](https://docs.openvino.ai/2023.0/openvino_docs_MO_DG_Deep_Learning_Model_Optimizer_DevGuide.html) on RHOCP using data center GPU cards or Intel Xeon processors. 6 | * Deployment mode – [OpenVINO Model Sever](https://github.com/openvinotoolkit/model_server) (OVMS) can be used to deploy the inference workloads in data center and edge computing environments on RHOCP. 7 | 8 | ## Prerequisites 9 | * Provisioned RHOCP cluster. Follow steps [here](/README.md#provisioning-rhocp-cluster) 10 | * Provisioning Intel Data Center GPU Flex Series. Follow steps [here](/README.md#provisioning-intel-hardware-features-on-rhocp) 11 | * Setup node feature discovery (NFD). Follow the steps [here](/nfd/README.md) 12 | * Setup out of tree drivers for Intel GPU provisioning. Follow the steps [here](/kmmo/README.md) 13 | * Setup Intel device plugins operator and create Intel GPU device plugin. Follow the steps [here](/device_plugins/README.md) 14 | 15 | ## Install RHOAI 16 | The Red Hat certified RHOAI operator is published at [Red Hat Ecosystem Catalog](https://catalog.redhat.com/software/container-stacks/detail/63b85b573112fe5a95ee9a3a). You can use the command line interface (CLI) or web console to install it. 17 | ### Install using CLI (To be added) 18 | ### Install using Web Console 19 | 1. On the RHOCP web console, click Operators → OperatorHub. 20 | 2. Search RedHat OpenShift AI Operator and click Install. The operator is installed in the namespace `redhat-ods-operator`. 21 | ### Verification 22 | 1. Navigate to Operators → Installed Operators page. 23 | 2. Ensure that in the redhat-ods-operator namespace, RedHat OpenShift AI status is InstallSucceeded 24 | 3. Click on `Search` -> `Routes` -> `rhods-dashboard` from the web console and access the RHOAI UI link. 25 | **Note:** When installing the operator, the default `kfdef` Custom Resource (CR) is created. This CR enables the dashboard for users to browse and launch Jupyter Notebooks projects on an RHOCP cluster. Please refer to this [link](https://github.com/red-hat-data-services/odh-deployer) for more details about `kfdef`. 26 | ## Install OpenVINO Operator 27 | The OpenVINO operator is published at [Red Hat Ecosystem Catalog](https://catalog.redhat.com/software/container-stacks/detail/60649a56209af65d24b7ca9e). You can use the CLI or web console to install it. 28 | ### Install using CLI (To be added) 29 | ### Install using Web Console 30 | Follow this [link](https://github.com/openvinotoolkit/operator/blob/v1.1.0/docs/operator_installation.md#operator-instalation) to install the operator via the web console. 31 | 32 | ## Work with Interactive Mode 33 | To enable the interactive mode, the OpenVINO notebook CR needs to be created and integrated with RHOAI. 34 | 1. Click on the `create Notebook` option from the web console and follow these [steps](https://github.com/openvinotoolkit/operator/blob/main/docs/notebook_in_rhods.md#integration-with-openshift-data-science-and-open-data-hub) to create the notebook CR. 35 | 2. Enable Intel Data Center GPU on RHOAI Dashboard- **Technical Preview feature** 36 | 37 | Create `AcceleratorProfile` in the `redhat-ods-applications` namespace 38 | ``` 39 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/e2e/inference/accelerator_profile_flex140.yaml 40 | ``` 41 | 42 | 3. Navigate to `openvino-notebooks` ImageStream and add the above created `AcceleratorProfile` key to the annotation field, as shown in the image below: 43 | 44 | ![Alt text](/docs/images/openvino-accelerator-field.png) 45 | 46 | 4. Navigate to `Search` -> `Networking` -> `Routes` from the web console and access `rhods-dashboard` route in the `redhat-ods-applications` namespace, as in the image below. Click on the location link to launch RHOAI dashboard. 47 | 48 | ![Alt text](/docs/images/rhods-dashboard-route.png) 49 | 50 | 5. If step 2 is successful, ```Intel® Data Center GPU Flex Series 140 ``` is shown in the accelerator dropdown menu in ```rhods-dashboard```. Users can run OpenVINO notebook image with Intel® Data Center GPU Flex Series 140 card. 51 | 52 | ![Alt text](/docs/images/accelerator-profile-dropdown.png) 53 | 54 | Follow the [link](https://github.com/openvinotoolkit/operator/blob/main/docs/notebook_in_rhods.md#integration-with-openshift-data-science-and-open-data-hub) for more details on the available Jupyter Notebooks. 55 | 56 | ## Work with Deployment Mode 57 | 1. From the web console, click on the ModelServer option in this [link](https://github.com/openvinotoolkit/operator/blob/v1.1.0/docs/operator_installation.md#operator-instalation) and follow the [steps](https://github.com/openvinotoolkit/operator/blob/v1.1.0/docs/modelserver.md#managing-model-servers-via-operator) to start the OVMS instance. 58 | 2. To enable the Intel Data Center GPU, make sure to modify the OVMS instance options according to the screenshot below. 59 | 60 | * Below images show `gpu.intel.com/i915` resource requests and limits for OVMS 61 | 62 | ![Alt text](/docs/images/Ovms-Gpu-resource-limit.png) 63 | 64 | ![Alt text](/docs/images/Ovms-Gpu-resource-request.png) 65 | 66 | # Enable Intel Gaudi AI Accelerator with RHOAI 67 | 68 | * From web console: 69 | 70 | To enable and use the Intel Gaudi accelerator on RHOAI web console, follow the [documentation](https://docs.redhat.com/en/documentation/red_hat_openshift_ai_self-managed/2.11/html/working_with_accelerators/intel-gaudi-ai-accelerator-integration_accelerators#enabling-intel-gaudi-ai-accelerators_accelerators). 71 | * From cli: 72 | 73 | Deploy the ```accelerator_profile_gaudi.yaml``` in the redhat-ods-applications namespace. 74 | 75 | ``` 76 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/e2e/inference/accelerator_profile_gaudi.yaml 77 | ``` 78 | ## Intel Gaudi AI Accelerator with Intel® Gaudi AI Software Tools Containers on OpenShift AI 79 | To use Intel Gaudi AI Accelerator with Intel® Gaudi AI Software Tools Containers on OpenShift AI, follow the [documentation](https://github.com/intel/ai-containers/blob/main/enterprise/redhat/openshift-ai/gaudi/README.md). 80 | 81 | ## See Also 82 | [GPU accelerated demo with OpenVINO](https://www.youtube.com/watch?v=3fTz_k4JT2A) -------------------------------------------------------------------------------- /one_click/gaudi_provisioning_playbook.yaml: -------------------------------------------------------------------------------- 1 | # Copyright (c) 2024 Intel Corporation 2 | # SPDX-License-Identifier: Apache-2.0 3 | - hosts: localhost 4 | vars: 5 | kubeconfig_path: "~/.kube/config" 6 | environment: 7 | KUBECONFIG: "{{ kubeconfig_path }}" 8 | tasks: 9 | - name: Install Dependencies 10 | tags: 11 | - install_dependencies 12 | block: 13 | - name: NFD - Install Node Feature Discovery Operator 14 | tags: 15 | - nfd 16 | block: 17 | - name: NFD - Create openshift-nfd namespace 18 | k8s: 19 | name: openshift-nfd 20 | api_version: v1 21 | kind: Namespace 22 | state: present 23 | wait: yes 24 | - name: NFD - Create an nfd-operator group v1 25 | k8s: 26 | definition: 27 | apiVersion: operators.coreos.com/v1 28 | kind: OperatorGroup 29 | metadata: 30 | generateName: openshift-nfd- 31 | name: openshift-nfd 32 | namespace: openshift-nfd 33 | spec: 34 | targetNamespaces: 35 | - openshift-nfd 36 | wait: yes 37 | - name: NFD - Create subscription for RH NFD operator 38 | k8s: 39 | definition: 40 | apiVersion: operators.coreos.com/v1alpha1 41 | kind: Subscription 42 | metadata: 43 | name: nfd 44 | namespace: openshift-nfd 45 | spec: 46 | channel: "stable" 47 | installPlanApproval: Automatic 48 | name: nfd 49 | source: redhat-operators 50 | sourceNamespace: openshift-marketplace 51 | wait: yes 52 | wait_condition: 53 | reason: AllCatalogSourcesHealthy 54 | type: CatalogSourcesUnhealthy 55 | status: 'False' 56 | - name: NFD - Wait until the nfd-operator-controller Deployment is available 57 | k8s_info: 58 | kind: Deployment 59 | wait: yes 60 | name: nfd-controller-manager 61 | namespace: openshift-nfd 62 | wait_condition: 63 | type: Available 64 | status: 'True' 65 | reason: MinimumReplicasAvailable 66 | - name: KMM - Install Kernel Module Management Operator 67 | tags: 68 | - kmm 69 | block: 70 | - name: KMM - Create openshift-kmm namespace 71 | k8s: 72 | name: openshift-kmm 73 | api_version: v1 74 | kind: Namespace 75 | state: present 76 | wait: yes 77 | - name: KMM - Create OperatorGroup v1 in openshift-kmm namespace 78 | k8s: 79 | definition: 80 | apiVersion: operators.coreos.com/v1 81 | kind: OperatorGroup 82 | metadata: 83 | name: kernel-module-management 84 | namespace: openshift-kmm 85 | wait: yes 86 | - name: KMM - Create Subscription for KMM Operator 87 | k8s: 88 | definition: 89 | apiVersion: operators.coreos.com/v1alpha1 90 | kind: Subscription 91 | metadata: 92 | name: kernel-module-management 93 | namespace: openshift-kmm 94 | spec: 95 | channel: stable 96 | installPlanApproval: Automatic 97 | name: kernel-module-management 98 | source: redhat-operators 99 | sourceNamespace: openshift-marketplace 100 | wait: yes 101 | wait_condition: 102 | reason: AllCatalogSourcesHealthy 103 | type: CatalogSourcesUnhealthy 104 | status: 'False' 105 | - name: KMM - Wait until the kmm-operator-controller Deployment is available 106 | k8s_info: 107 | kind: Deployment 108 | wait: yes 109 | name: kmm-operator-controller 110 | namespace: openshift-kmm 111 | wait_condition: 112 | type: Available 113 | status: 'True' 114 | reason: MinimumReplicasAvailable 115 | - name: Install Intel Gaudi Base Operator 116 | tags: 117 | - intel-gaudi 118 | block: 119 | - name: Install Intel Gaudi Base Operator 120 | k8s: 121 | state: present 122 | src: "../gaudi/gaudi_install_operator.yaml" 123 | wait: yes 124 | - name: Wait until the Intel Gaudi controller-manager Deployment is available 125 | k8s_info: 126 | kind: Deployment 127 | wait: yes 128 | name: controller-manager 129 | namespace: habana-ai-operator 130 | wait_condition: 131 | type: Available 132 | status: 'True' 133 | reason: MinimumReplicasAvailable 134 | - name: NFD - Install NFD CRs 135 | block: 136 | - name: NFD - Create NFD discovery instance for Intel Gaudi 137 | k8s: 138 | state: present 139 | src: "../gaudi/gaudi_nfd_instance_openshift.yaml" 140 | wait: yes 141 | - name: Install Intel Gaudi DeviceConfig CR 142 | block: 143 | - name: Create Intel Gaudi DeviceConfig 144 | k8s: 145 | state: present 146 | src: "../gaudi/gaudi_device_config.yaml" 147 | wait: yes 148 | - name: Verify Intel Gaudi Resources 149 | tags: 150 | - gaudi_resource_test 151 | block: 152 | - name: Get Gaudi Node Resource Information 153 | kubernetes.core.k8s_info: 154 | api: v1 155 | kind: Node 156 | label_selectors: 157 | - "kmm.node.kubernetes.io/habana-ai-operator.intel-gaudi-module.device-plugin-ready=" 158 | - "kmm.node.kubernetes.io/habana-ai-operator.intel-gaudi-module.ready=" 159 | wait: yes 160 | wait_timeout: 120 161 | register: cluster_nodes_info 162 | until: 163 | - cluster_nodes_info.resources is defined 164 | - name: Print cluster resources 165 | debug: 166 | msg: 167 | - "Please verify Capacity and Allocatable Habana Gaudi Resources on the node - " 168 | - "Capacity: { 169 | 'habana.ai/gaudi': {{ cluster_nodes_info.resources[0].status.capacity['habana.ai/gaudi'] }}," 170 | - "Allocatable Resources: { 171 | 'habana.ai/gaudi': {{ cluster_nodes_info.resources[0].status.allocatable['habana.ai/gaudi'] }}," -------------------------------------------------------------------------------- /device_plugins/deploy_qat.md: -------------------------------------------------------------------------------- 1 | # Create Intel QAT Device Plugin CR 2 | 3 | ## Create a CR via web console 4 | 1. Go to **Operator** -> **Installed Operators**. 5 | 2. Open **Intel Device Plugins Operator**. 6 | 3. Navigate to tab **Intel QuickAssist Technology Device Plugin**. 7 | 4. Click **Create QatDevicePlugin** -> set correct parameters -> Click **Create** 8 | 5. Optional: If you want to make any customizations, select YAML view and edit the details. When you are done, click **Create**. 9 | 10 | ## Verify via web console 11 | 1. Verify CR by checking the status of **Workloads** -> **DaemonSet** -> **intel-qat-plugin**. 12 | 2. Now `QatDevicePlugin` is created. 13 | 14 | ## Create CR via CLI 15 | Apply the CR yaml file: 16 | ``` 17 | $ oc apply -f https://raw.githubusercontent.com/intel/intel-technology-enabling-for-openshift/main/device_plugins/qat_device_plugin.yaml 18 | ``` 19 | 20 | ## Verify via CLI 21 | Verify that the device plugin CR is ready: 22 | ``` 23 | $ oc get QatDevicePlugin 24 | ``` 25 | Output: 26 | ``` 27 | NAME DESIRED READY NODE SELECTOR AGE 28 | qatdeviceplugin-sample 1 1 {"intel.feature.node.kubernetes.io/qat":"true"} 3h27m 29 | ``` 30 | 31 | # Verify QAT Device Plugin 32 | After the plugin is deployed, use below command to verify QAT resources: 33 | ``` 34 | $ oc describe node | grep qat.intel.com 35 | qat.intel.com/cy: 32 36 | qat.intel.com/cy: 32 37 | qat.intel.com/dc: 32 38 | qat.intel.com/dc: 32 39 | ``` 40 | **Note**: By default the device plugin registers half resources each for `qat.intel.com/cy` and `qat.intel.com/dc` respectively. For more details about the QAT resources configuration, please refer to the QAT Device Plugin Configuration section below. 41 | 42 | # QAT Device Plugin Configuration 43 | > **Note**: The QAT device plugin can be configured with the flags. In this release, only the configurations in the table below are verified and supported on RHOCP. 44 | 45 | For more details about the QAT device plugin configuration flags, see [Modes and Configurations Options](https://github.com/intel/intel-device-plugins-for-kubernetes/blob/main/cmd/qat_plugin/README.md#modes-and-configuration-options). 46 | 47 | | Flag | Configuration | Description | 48 | | ---- | ---- | ---- | 49 | | `-dpdk-driver` | vfio-pci | Using vfio-pci driver to manage QAT VFIO device. See details [here](https://doc.dpdk.org/guides/linux_gsg/linux_drivers.html) | 50 | | `-kernel-vf-drivers` | 4xxxvf | Supporting 4xxx QAT device
**Note**: Verified on 4th Gen Intel® Xeon® Scalable processors. See details [here](https://github.com/intel/qatlib/blob/main/INSTALL#L72) | 51 | | `-max-num-devices ` | 128 | It is the maximum VF device it can support for 4xxx QAT device. If the number exceeds the maximum number the QAT device supports, then the maximum number will be enabled. | 52 | | `-provisioning-config ` | Name of ConfigMap | See section [QAT resource configuration](/device_plugins/deploy_qat.md#qat-resource-configuration-experimental) | 53 | 54 | ## QAT Resource Configuration (experimental) 55 | 56 | **NOTE**: In this release, this is an experimental feature. The efforts to [enhance this feature](https://github.com/intel/intel-device-plugins-for-kubernetes/issues/1529) and [make it more stable](https://github.com/intel/intel-device-plugins-for-kubernetes/issues/1542) are on going. 57 | 58 | Users can use the steps below to customize the QAT resource configuration: 59 | 1. Create the configmap for qat resource configuration 60 | ``` 61 | $ oc create configmap --namespace=openshift-operators --from-literal "qat.conf=ServicesEnabled=