Kunpeng Health Inspector Functions

The Kunpeng Health Inspector can collect hardware information such as CPU, memory, NIC, and PCIe data in a lightweight manner, generate health reports and tuning suggestions, and help detect performance deterioration caused by hardware faults or configuration errors. It supports cross-report comparison and comparison against custom baseline data, with interactive scenario-based parameter configuration. It can run on servers powered by the Kunpeng 920 and 950 processors, as well as on certain servers equipped with the Kunpeng 920 new model.

Prerequisites

The Kunpeng Health Inspector has been installed.
The tool is contained in an independent package (devkit-kspect-x.x.x-Linux-aarch64.tar.gz). Extract the package and switch to the tool directory.

Command Function

The Kunpeng Health Inspector is a lightweight and precise tool for collecting static Kunpeng hardware information. It swiftly collects data on server hardware, such as CPUs, memory, network, storage, PCIe, virtual machines (VMs), sensors, software, and module dependencies, and offers performance tuning suggestions based on the collected data.

These functions are available on Kunpeng hardware and are supported on physical machines, VMs, and containers. If certain data is not supported, a message is displayed. Hardware information obtained from VMs and containers may be inconsistent with that from physical machines.
The tool can run on openEuler, CentOS, Debian, and Ubuntu. The OS version must be the same as that supported by the DevKit. For a non-openEuler OS, some fields may be missing when server hardware information is collected.
OSs with kernel versions 4.19, 5.10, 6.6, and 5.15 are supported.

Syntax

./kspect [-h] [-l {0,1,2,3,4}] [-lang {0,1}] [--remote] [-s] [-c COMMAND]{system,os,bios,software,cpu,numa,memory,network,storage,pcie,ascend,ub,bmc,config,cluster,all,report,domain} ...

Example

./kspect -lang 0 -h

Command output:

NAME
    Kunpeng Health Inspector

USAGE
    kspect [-h] [-l {0,1,2,3,4}] [-lang {0,1}] [--remote] [-s] [-c COMMAND] {system,os,bios,software,cpu,numa,memory,network,storage,pcie,ascend,ub,bmc,config,cluster,all,report,domain} ...

POSITIONAL ARGUMENTS    
    {system,os,bios,software,cpu,numa,memory,network,storage,pcie,ascend,ub,bmc,config,cluster,all,report,domain}
    Select health check items.

    system
    Displays all system information, including the server, vendor, and mainboard information.

    os
    Displays the OS name, system version, kernel version, and boot parameters.

    bios
    Displays BIOS information, including the boot type, cache mode, and PCIe rate.
    Note: To use this feature, you need to configure the BMC login information (including the user name and password).

    software
    Displays software information, including GCC, glibc, Binutils, Python, KVM, and Docker.

    cpu
    Displays CPU information, including the CPU model and list.

    numa
    Displays NUMA information, including CPU, memory, PCIe, network, and NVMe information.

    memory
    Displays memory information, including common memory information and DIMM information.

    network
    Displays network information, including NIC and IRQ information.

    storage
    Displays storage information, including drives and partitions.

    pcie
    Displays PCIe information, including the PCIe driver and IRQ information.

    ascend
    Displays Ascend NPU information, including npu-smi and ascend-dmi.

    ub
    Collects UnifiedBus data.

    bmc
    Displays BMC usage information. To use this feature, you must have the permission to use ipmitool or configure BMC login information (including the BMC IP address, user name, and password).

    config
    Checks and displays the dependency between modules.

    cluster
    Collects the remote cluster data and sends back reports.

    all
    Displays full system information.
    Note: To collect all information, you must provide the BMC login information (IP address, user name, and password). If you do not provide this information, ipmitool is used instead, and the report does not include information from modules that have strong dependencies on the BMC. To protect your account, do not enter your password in public places or save it on other people's devices or public devices. Ensure that your network environment is secure.

    report
    Displays historical reports or analyzes the differences between reports.

    domain
    Configures scenario parameters.

OPTIONS
    -h, --help
    Obtains help information and exits.

    -l {0,1,2,3,4}, --log-level {0,1,2,3,4}
    Specifies the log level, which is 0(debug) | 1(info) | 2(warning) | 3(error) | 4(notset). The default is 1(info).

    -lang {0,1}, --language {0,1}
    Sets the display language of the tool. The options are 1 (Chinese, default) and 0 (English).
    Note: If the current environment does not support the selected language, 0 is used.

    --remote
    Indicates whether to collect remote node information.
    Note: The tool will be transferred to the /tmp directory on the remote node. After the collection is complete, the tool is deleted. To use the tool on the remote node, you need to obtain the SSH login information of the remote node, including the IP address and port (xxx.xxx.xxx.xxx:Port, with the porting defaulting to 22), username, and password.

    -s, --skip-bmc
    Skips BMC collection.

    -c COMMAND, --bmc-command COMMAND
    Runs the user-defined BMC information collection command.

Parent topic: Kunpeng Health Inspector (CLI)