Install Alauda AI Infrastructure Accelerator Virtualization
This page describes how to install the v2.10.0 HAMi Operator in ACP. HAMi-WebUI and the HAMi Ascend Device Plugin are separately packaged components with their own compatibility and installation requirements.
TOC
Before you beginChoose a deployment scenarioPrepare the Operator packageInstallAlauda AI Infrastructure Accelerator VirtualizationContinue with backend-specific componentsNext stepsBefore you begin
- Cluster administrator access to the target ACP cluster.
- The v2.10.0 Operator package supported by the target ACP and Kubernetes versions. See Versions and Components.
- Confirm that no earlier HAMi cluster plugin owns the scheduler, webhook, RuntimeClass, or cluster RBAC resources. Do not install both package forms in the same cluster.
- Accelerator nodes have the required vendor driver, runtime, and node readiness checks completed through the corresponding NVIDIA GPU vendor base documentation or Ascend NPU vendor base documentation.
- The target backend is supported by the current HAMi package and ACP release.
Choose a deployment scenario
Before installing HAMi, use Choose a HAMi Deployment Scenario to determine the required products, HAMi settings, installation sequence, and device exposure owner. Return to this page after selecting the scenario.
Select one vendor backend for the target cluster. This procedure does not cover enabling NVIDIA GPU and Ascend NPU backends in the same cluster.
Prepare the Operator package
Retrieve Alauda AI Infrastructure Accelerator Virtualization v2.10.0 from Alauda Cloud and make the Operator package available in ACP. Please contact Consumer Support for package access. For the standard upload procedure, see Upload Packages.
Install Alauda AI Infrastructure Accelerator Virtualization
-
Label the nodes that should run HAMi components. Use the node selector required by the selected backend and the
Hamicustom resource.For common NVIDIA backend examples:
For common Ascend backend examples:
WARNINGKeep one device exposure and resource allocation owner for the same physical devices.
If HAMi exposes devices on a node, do not also let a vendor device plugin or DRA driver expose the same devices on that node. For Ascend, use Switch the Ascend Device Plugin Owner for the maintenance procedure.
-
In ACP, go to
Administrator->Marketplace->OperatorHub, switch to the target cluster, and installAlauda AI Infrastructure Accelerator Virtualizationinhami-system. Approve the InstallPlan if the Subscription uses manual approval. Wait forhami-operator.v2.10.0to reachSucceededand for the Operator Deployment to become Ready. -
Create one
Hamicustom resource namedhamiinhami-system. Configure its values for the selected backend and runtime. The Operator manages the HAMi Helm release from this resource; do not runhelm installfor the same release separately.NOTEOn immutable OS nodes, keep the host and container hook paths aligned with the target runtime. Verify a GPU workload's actual memory limit after installation; a Ready device-plugin Pod alone does not prove memory interception.
For overcommitment behavior, see Configure Overcommitment Ratio. For CDI preparation and driver paths, continue in the NVIDIA GPU vendor base documentation.
-
Confirm that the
Hamicustom resource reportsDeployed, then verify that the HAMi scheduler is running:For the NVIDIA scenario, also verify the built-in NVIDIA device plugin:
-
For the NVIDIA scenario, confirm that nodes expose the expected HAMi NVIDIA resource keys:
For an Ascend scenario, install
Alauda Build of HAMi Ascend Device Pluginbefore checking node allocatable resources. The core HAMi installation does not expose Ascend devices by itself. -
For the NVIDIA scenario, register ACP quota metadata for the HAMi resource names that users will request. For an Ascend scenario, complete the HAMi Ascend Device Plugin installation and confirm its resource names before registering the metadata. For the ACP-owned resource metadata procedure, see Accelerator Resource Quota.
Continue with backend-specific components
- For HAMi-managed Ascend allocation or vNPU, continue with Install HAMi Ascend Device Plugin.
- For the optional HAMi resource WebUI, continue with Install Alauda AI Infrastructure Accelerator Virtualization Console.
Direct Ascend NPU allocation without HAMi scheduling is outside this HAMi-managed installation path. Continue with Request Ascend NPU resources directly for that workload model.