文档中心API Explorer
ZHEN
文档中心›ZVF›ZStack ZSphere›用户手册ZVF

Host

A host is a sub-resource of a cluster and can run one or more virtual machines. This section describes how to use hosts in the following chapters:

Host Basic Operations

You can understand the basic operations supported by hosts from the perspective of adding, deleting, modifying, and querying.

Add a Host

The platform provides multiple entry points to add hosts. You can add one or multiple hosts from the following two main entry points:
  • In the navigation bar on the left side of the platform page, right-click the target cluster and click Add Host.
  • In the navigation bar on the left side of the platform page, select the target cluster. Then, on the right side of the platform page, click Actions > Add Host, or in the Hosts sub-page, click Add Host.

ZStack ZSphere supports configuration of the following three major categories of information:

Basic Information: includes name description, associated cluster, and labels
  • Name: host name
  • Description: host description
  • Cluster: cluster where the host resides
  • Tag: supports binding one or more labels to identify different hosts. For more details, see Tag Management
Host Information: includes addition method, IP address range, and SSH configuration:
  • Addition Method: add a single host or multiple hosts. When adding multiple hosts, ensure all hosts have the same SSH configuration.
  • IP Address: enter the IP address or IP address range of the host based on the addition method.
  • SSH Port: SSH port of the host, default is 22.
  • SSH Username: username for the host. Default is root.
  • SSH Password: password for the SSH username.
Other Information Configuration: includes IOMMU enablement status and Intel EPT hardware assistance
  • Scan Host IOMMU Setting: whether the IOMMU (Input/Output Memory Management Unit) function is enabled, used for passthrough of external devices and virtualization scenarios. By default, it is disabled on x86 architecture and enabled on ARM architecture where the BIOS IOMMU is called SMMU.
    Note: Before enabling, make sure the IOMMU option is enabled in the host's BIOS.
  • Intel EPT Hardware Assist: whether Intel EPT hardware assistance is enabled on Intel CPUs to improve CPU performance. Default is enabled.

After clicking OK, the creation process is complete.

After the first host is added to the cluster, ZStack ZSphere automatically creates a default distributed switch, default distributed port group, and default Kernel adapter based on the related configurations of this host for centralized management of the host's management network. For more information, see Network Resource.

Modify a Host

If you need to modify the name or description of an existing host, on the target host page, click ActionsEdit Name and Description, and modify the corresponding information in the pop-up window.

If you need to modify the IOMMU enablement status and Intel EPT hardware assistance settings of an existing host, on the target host page, click ActionsModify Configuration, and modify the corresponding information in the pop-up window.

Access a Host

You can access the host system through the following three methods:
  • Webshell Terminal Access: you can directly access the host system by clicking on the small terminal window of the target host, or by clicking Actions > Enter Web Terminal.
  • SSH Access: you can log in using remote login software by entering the SSH information entered when adding the host. To modify SSH information, click Actions > Update SSH Information.
  • IPMI Access: if you have managed the host through IPMI, you can access the host using IPMI management software. To modify IPMI information, click Actions > Update IPMI Information.

View a Host

If you need to check the usage and allocation of CPU and memory resources for the host, as well as the usage and allocation of all virtual machines under the host, go to the host's Overview details page. For more information, see Capacity Monitoring.

If you need to check the usage trends of CPU, memory, disk, and NIC resources for the host and its virtual machines over time, go to the host's Monitoring tab. For more information, see Resource Performance Monitoring.

Delete a Host

If you need to delete an existing host, on the target host page, click ActionsDelete, and the host will be deleted. You can also delete hosts in batches on the data center resources Cluster and Host page or the cluster resources Host page.
Note:
  • If the cluster to which the host belongs has loaded shared storage, this operation will stop all VMs on the host. VMs with high availability enabled will automatically migrate to other hosts within the cluster with sufficient resources and restart.
  • If the cluster to which the host belongs has loaded local storage, this operation will delete all VMs and disks on the host. Proceed with caution.

Host Hardware Device

After adding the host to the ZStack ZSphere platform, you can enter the host's Hardware Device page to view and manage the host's hardware and devices:

Host NUMA Topology

Host NUMA Topology: A pNUMA topology (physical NUMA topology) is the topology of the host NUMA nodes predefined by the CPU vendor based on the host NUMA architecture.

Definitions

  • Non-Uniform Memory Access (NUMA): Non-uniform memory access (NUMA) is a computer memory design where the memory access time depends on the memory location relative to the CPU. Under NUMA, a processor can access its own local memory faster than non-local memory and thus improves VM performance.
  • pNUMA Node: A pNUMA node (physical NUMA node) is a host NUMA node predefined based on the host NUMA architecture. It is used to manage the CPUs and memory of the host. A host can have one or more pNUMA nodes. A pNUMA node primarily consists of one or more physical CPU cores (pCPU) and local memory.
  • vNUMA Node: A vNUMA node (virtual NUMA node) is generated by passing-through associated pNUMA nodes via CPU pinning. It is used to manage the CPUs and memory of a virtual machine. A vNUMA node primarily consists of one or more virtual CPU cores (vCPU) and local memory.
  • vNUMA Topology: A vNUMA topology (virtual NUMA topology) is the topology of VM NUMA nodes generated by passing-through associated pNUMA nodes via CPU pinning.
  • Local Memory: Local memory is the memory that a CPU (pCPU or vCPU) accesses through the Uncore iMC (Integrated Memory Controller) of the same NUMA (pNUMA or vNUMA) node. Compared with accessing non-local memory, accessing local memory has lower latencies.

Functionality Principle

After adding the host, ZStack ZSphere supports viewing the host's pNUMA topology and configuring vNUMA for virtual machines running on the host based on this topology.

ZStack ZSphere Virtual machine vNUMA configuration is achieved through CPU pinning, which strictly associates the virtual machine's vCPU with the host's pCPU, allocating specific pCPUs to the virtual machine. During vNUMA configuration, all vCPUs of the virtual machine are pinned to pCPUs, and each vCPU's pinned pCPUs are located within the same pNUMA node.

After vNUMA configuration, the virtual machine directly passes through the associated host pNUMA node topology, generating one or more vNUMA nodes that form the virtual machine's vNUMA topology. Virtual machine vCPUs prioritize accessing local memory within the same node based on the vNUMA topology.

pNUMA Topology

Go to the target host's Overview, and click View pNUMA Topology in the hardware overview information box. ZStack ZSphere host pNUMA topology information is as follows:

  • Displays all pNUMA nodes of the host and the virtual machine information associated with each node.
  • Total memory is the local memory of the pNUMA node that can be directly accessed by the pCPU.
  • Free memory is the local free memory of the pNUMA node that can be directly accessed by the pCPU.
  • Both total memory and free memory are based on the actual hardware physical memory capacity of the pNUMA node.

LUN

On the host Hardware Device tab, click LUN to check the block device on the host.

View SCSI Devices

You can view basic information and path information about SCSI block devices scanned on the host.

Procedure:
  1. Navigate to Inventory > Host and VM.
  2. Select the target host, then click the host name to enter the details page.
  3. Click Hardware Device > LUN > SCSI Device.

    On the SCSI Device tab, view the supplier, model, capacity, WWN, WWID, number of mounted virtual machines, type, and source information for the block devices.

  4. Click the name of the target block device, then click Paths to view all available paths for the block device and the status of each path.

View NVMe Devices

You can view NVMe LUNs connected to the host via network protocols. Local NVMe PCIe disks on the host can be viewed in the Physical Disk list.

Procedure:
  1. Navigate to Inventory > Host and VM.
  2. Select the target host, then click the host name to enter the details page.
  3. Click Hardware Device > LUN > NVMe Devices.

    On the NVMe Devices tab, view the supplier, model, WWN, capacity, WWID, and type information for the block devices.

Host NIC

On the host Hardware Device tab, click Physical NIC to check and manage the physical NICs and bonds on the host.

Physical NIC - Standard Configuration Changes

You can perform editing or IP address modification operations on physical NICs of the host:
  • Edit: Modify the description of the physical NIC.
  • Modify IP Address: If the NIC has not been added to a bond interface and is not part of a distributed switch, you can modify the IP address and subnet mask of the NIC as needed.

Physical NIC - SR-IOV Virtualization

You can virtualize and partition a single physical NIC into multiple VF (Virtual Function) NICs based on the SR-IOV specification, which can then be directly assigned to virtual machines. This allows for near-native I/O performance and reduces consumption of host CPU resources.

Prerequisites
  • Ensure that the physical NIC supports SR-IOV partitioning.
  • Ensure that the BIOS of the host containing this physical NIC has Intel VT-d/AMD IOMMU and SR-IOV features enabled.
  • Ensure that the IOMMU readiness status of the host containing this physical NIC is Available.
Procedure
  1. Navigate to Host Details Page > Hardware Device > Physical NIC.
  2. Select the target physical NIC, then click Actions > Configure SR-IOV.
  3. In the Configure SR-IOV dialog, enable the SR-IOV Status, and specify the number of VF NICs to be created.
Notes
  • If the physical NIC is already configured in a bond, continuing to use the SR-IOV feature may affect communication between VF cards and vNIC cards. It is recommended to first perform SR-IOV partitioning on the physical NIC, then configure the bond from the distributed switch, selecting only one partitioned physical NIC per host.
  • If VF cards are currently being used by virtual machines, turning off the SR-IOV Status switch will simultaneously unload the related NICs from the virtual machines.
  • Virtual machines that are powered on and have loaded VF cards do not support migration operations. You need to power off the virtual machine or first unload the VF cards before migration can occur.

Physical NIC - View and Maintain LLDP Information

View Peer Device Information

Using the Link Layer Discovery Protocol (LLDP), you can identify the physical switch port of the distributed switch to which the physical NIC is connected. Before viewing the peer device information, ensure that your NIC supports LLDP and that the peer switch device has LLDP enabled.

  • Recommended Hardware Specifications:

    • Switches: Huawei Switches, H3C Switches, and Shengke Switches.
    • NICs: Intel 82599ES, Intel x710, Intel x722, and Mellanox CX4.

    Supported OS Types:

    • x86: H84r or x86_KylinV10P3
    • ARM: arm_KylinV10P3 or H22e

  • Procedure:

    1. Click on the physical NIC name to open the details page.
    2. Click LLDP to view the peer device information. For the interpretation of LLDP TLV units, refer to 表 1.

Modify LLDP Mode

Using the Link Layer Discovery Protocol (LLDP), you can obtain peer device information connected to the port or send your own device information to neighboring devices directly connected to you, allowing for link communication status queries and judgments.

Steps:
  1. Click on the physical NIC name to open the details page.
  2. In the NIC details page, click LLDP.
  3. On the LLDP tab, click Modify.
  4. In the Modify NIC LLDP Mode dialog, select the LLDP mode from the dropdown options.
    • Receive Only: The default mode, which only parses and displays the peer LLDP information received on this port.
    • Send Only: Sends the LLDP information from this port but does not parse the received LLDP information. Peer device information cannot be viewed in this mode.
    • Send and Receive: Parses and displays the peer LLDP information received on this port while also sending the LLDP information from this port to the connected peer device.
    • Disabled: Does not parse received LLDP information and does not send any LLDP information from this port. Peer device information cannot be viewed in this mode.
  5. Click OK.
Field Description
Device ID Chassis ID, which is the bridge MAC address of the sending device.
Port ID Port ID, which identifies the port.
Management Address Management Address, which is the management address of the sending port.
TTL Time to Live, indicating how long this device's information remains on neighbor devices.
Port Description Port Description, which provides details about the port.
System Name System Name, which identifies the name of the device.
System Description System Description, providing information about the system.
System Capabilities System Capabilities, indicating the primary functions of the system and those that are in use.
VLAN ID Primary VLAN ID of the port.
Aggregation Status Link Aggregation, indicating whether the port supports link aggregation and if it has been enabled.
MTU Maximum Frame Size, which is the maximum frame length supported by the port, taken from the configured Maximum Transmission Unit (MTU) value.

Bond - View and Configure

If you associate a physical network port with a distributed switch for aggregation when creating the distributed switch, you can view the aggregation port on the corresponding host resource's physical NIC Bond page. Click the Refresh button to view the latest information, including the port's aggregation mode, aggregation port status, rate, associated distributed switch, IPv4 address, and creation time. If you need to manage the aggregation port configuration for this host, refer to Manage Joined Hosts.

GPU Device

On the Hardware Device tab of the host, select the Physical GPU Device or vGPU Device tab to view and manage physical GPU devices and vGPU devices for that host.

Physical GPU Device

You can perform different operations on physical GPU devices based on various scenarios:
  • Enable and Disable Scenarios: Enable & Disable
    • If you want to pass through the physical GPU device directly to a virtual machine, ensure that the device is enabled. Click the Enable button to enable it.
    • If you no longer wish to pass through the physical GPU device to any virtual machines, click the Disable button to disable it.
      Note: After disabling, the physical GPU device currently in use by a virtual machine will continue to function normally until it is uninstalled.
  • Virtualization Scenarios: Virtualization Partitioning & Restoration
    • Virtualization Partitioning: Partition an unpassed-through physical GPU device into vGPU devices of specified specifications. Before partitioning, ensure the following conditions are met:
      • The physical GPU model supports virtualization partitioning.
      • The physical GPU is not passed through to any virtual machine.
      • The host BIOS has Intel VT-d / AMD IOMMU enabled, and the host kernel has IOMMU support enabled.
      • The host's IOMMU readiness status in the platform is set to Available.
      Different manufacturers have slightly different methods for partitioning physical GPUs virtually:
      • NVIDIA: Supports partitioning NVIDIA physical GPUs according to selected specifications individually.
      • AMD: Supports partitioning all AMD physical GPUs on the current host simultaneously based on the selected quantity.
    • Virtualization Restoration: Restore vGPU devices back to physical GPU devices. Before restoration, ensure that all vGPUs created from this physical GPU have been uninstalled from virtual machines. The method of restoring physical GPUs varies by manufacturer:
      • NVIDIA: Ensure that all vGPUs related to this NVIDIA physical GPU have been uninstalled from virtual machines before restoration.
      • AMD: Ensure that all AMD vGPUs on the current host have been uninstalled from virtual machines before restoring AMD vGPUs.

vGPU Devices

You can perform different operations on vGPU devices based on various scenarios:
  • Enable and Disable vGPU Device Scenarios: Enable & Disable
    • If you want to pass through the vGPU device directly to a virtual machine, ensure that the device is enabled. Click the Enable button to enable it.
    • If you no longer wish to pass through the vGPU device to any other virtual machines, click the Disable button to disable it. After disabling, the vGPU device currently in use by a virtual machine will continue to function normally until it is uninstalled.

USB Device

On the host Hardware Device page, select the USB Device tab to view and manage the USB devices for that host.

You can perform different operations on USB devices based on various scenarios:
  • Device Renaming Scenario: If you wish to rename a USB device to better align with your business needs, click the Edit Device Name button to rename it.
  • Enable and Disable Scenarios:
    • If you want to pass through a USB device to a virtual machine, ensure that the device is enabled. Click the Enable button to enable it.
    • If you no longer want to pass through the USB device to any other virtual machines, click the Disable button to disable it. After disabling, the USB device currently in use by a virtual machine will continue to function normally until it is uninstalled.
  • VM Attach/Detach Scenarios: Attach VM & Detach VM
    • Attach VM: Pass the USB device directly to a virtual machine, supporting both Direct Connection and Forwarding modes.
      • Direct Connection: Attach the USB device from the host where the VM resides to the VM. This USB device must be unloaded if the VM is migrated.
      • Forwarding: Attach the USB device from any host within the data center where the VM resides to the VM. This USB device does not need to be unloaded if the VM is migrated.
      When loading a USB device to a virtual machine, note the following:
      • The same USB device can only be passed through to one virtual machine at a time.
      • A single virtual machine supports up to 1 USB 1.0 device, up to 6 USB 2.0 devices, and up to 4 USB 3.0 devices.
      • A running VM or a VM with local storage in a stopped state only supports loading available USB devices from the host where the VM resides; cross-host USB device loading is not supported.
      • A VM stopped on shared storage supports loading multiple USB devices from any host within its cluster.
    • Detach VM: Detach the USB device from the virtual machine.
      Note: This action will interrupt read/write operations for the USB device, proceed with caution.

PCIe Device

On the host Hardware Device page, select the PCIe Device tab to view and manage the PCIe devices for that host.

You can enable passthrough functionality for PCIe devices on the host. Passthrough devices provide an effective way to use resources and improve environment performance.
  • Passthrough PCI Devices: Displays PCIe devices with passthrough functionality enabled, available for virtual machine use.
  • All PCI Devices: Displays all PCIe devices detected on the host. You can switch devices with a Configurable passthrough state to passthrough devices.
Note:
  • Before switching a PCIe device to passthrough, ensure that IOMMU is enabled on the host and that the IOMMU readiness state is available.
  • If a PCIe device is already loaded by a virtual machine, it cannot be switched to passthrough. Please unload the virtual machine before retrying.
上一篇Cluster下一篇Image Storage