No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search


To have a better experience, please upgrade your IE browser.


FusionStorage V100R006C20 Block Storage Service Software Installation Guide 07

Rate and give feedback:
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).
Installing a CVM

Installing a CVM


WhenFusionStorage Block is connected to VMware vSphere, a controller VM (CVM) is required to accommodate the FSA node because the FSA node cannot be installed on the ESXi host. If a server functions only as a storage node without ESXi installed, the FSA node can be installed directly on the server.

This topic instructs software commissioning engineers to install a VM created by VMware as a FusionStorage Block CVM. In the converged deployment mode, if a CVM has both the Virtual Block System (VBS) process and the Object Storage Device (OSD) process deployed, configure RAID controller cards and SSD cards to pass through to the CVM so that the OSD process can perform read and write operations to storage media.

Create and install a CVM on each ESXi host.



The requirements listed in System Requirements have been fulfilled.


No data is required for performing this operation.


FusionStorage Block


    Create a CVM.

    1. Use vClient to log in to the ESXi host and create a CVM.

      Table 6-6  CVM specifications
      Item Specifications
      Operating system (OS) SUSE Linux Enterprise Server 11 SP3 (64-bit)

      Number of CPUs for a CVM in separated deployment mode: 4

      Number of CPUs for a CVM in converged deployment mode: 8

      Memory (GB)

      See see Memory in System Requirements..

      NIC (Network adapter)
      The formula for calculating memory usage is as follows: Memory occupied by each server = Memory occupied by the operating system + Memory occupied by the VBS process.
      • For details about the memory occupied by the operating system, see the memory requirements in product documentation for corresponding products.
      • For details about the memory occupied by the VBS process, see the memory capacity of compute nodes in Memory in System Requirements. .
      Disk (GB)

      One 100 GB disk that uses local storage on the host

      If other options are required, apply the default specifications.

      The VM must be stopped after it is created, because other VM parameters need configuration.

    2. If multiple ESXi hosts are required, create a CVM on each ESXi host and set the CVM to start with the host startup.

    Configure the CVM.

    1. Right-click the CVM and choose Edit Setting to set CVM parameters on the VM attribute page.

      • CPU resource allocation: Allocate the CPU resources to the CVM to ensure sufficient hardware resources are reserved for the CVM even when service loads are heavy.

        • If the CPU quota is set to Normal, set the reservation value (on the lower part of the red box) to the actual CPU clock rate (on the upper part of the red box), as shown in Figure 6-6.

          Figure 6-6  CPU quota
        • If the hyper-threading core sharing mode is set to None, clear the configuration of Scheduling Affinity, as shown in Figure 6-7.

          When the VMware Distributed Resource Scheduler (DRS) feature is enabled for the cluster and its automation level is set to full automation, Scheduling Affinity will be no longer displayed, and you only need to set the hyper-threading core sharing mode to None.

          Users can right-click the cluster and choose Edit Setting to view the DRS status of the cluster.

          Figure 6-7  Hyper-threading core sharing
      • Memory resource allocation: Allocate the memory resources to the CVM to ensure sufficient hardware resources are reserved for the CVM even when service loads are heavy.

        Requirements: Reserve all the memory, as shown in Figure 6-8.
        Figure 6-8  Memory reservation

    2. Set the CVM parameters for all the other CVMs. For details, see 3.
    3. Power on all the CVMs.

    Install an OS.

    1. See Installing an OS on FSM to mount the operating system image file. See Installing an OS on Storage Nodes to install the operating system based on the site requirements.

      Ensure that the space in the following directories available after the OS is installed meets the running requirements of FusionStorage Block software.

      Directory Available space Description
      /opt 1 GB The required directory space must be provided for FusionStorage Block software to run properly.
      /tmp 1 GB The available space must be greater than 1 GB before installing the software. After the software is installed, you can clear the directory.
      /var/log 2 GB You are advised to create an independent partition for storing logs, avoiding adverse impacts caused by a large number of logs on OS running.

      Table 6-7 lists the minimum space requirements for system partitions. If your system disk space is sufficient, you are advised to properly expand the size of each partition based on the actual condition.

      Table 6-7  Server OS partition
      Mount Point


      Minimum Partition Size (GB) Recommended Partition Size (GB)
      / System root directory 20 20
      swap Swap partition 20 20
      /usr System program directory 5 20
      /opt Third-party software directory 5 40
      /tmp Directory for saving temporary files generated by users or during program running 5 40
      /var Directory for saving changed data during system running 5 5
      /var/log Log partition

      Other files cannot be saved in this partition.

      5 60

    Configure network information.

    The following operations describe how to configure network information for VMs running the SUSE Linux OS.

    1. In the remote control window of any CVM, log in to the CVM as user root.
    2. Run the following command to enter the path to the NIC configuration files:

      cd /etc/sysconfig/network

    3. Run the following command to query the NIC configuration files in the path and then make a note of the file names:


      Information similar to the following is displayed:



    4. Modify the NIC configuration files based on the network planes configured for the CVM network adapters.

      For example, if the management plane uses the first network adapter, the NIC configuration file name is ifcfg-eth0. If the storage plane uses the second network adapter, the NIC configuration file name is ifcfg-eth1. If the iSCSI plane uses the third network adapter, the NIC configuration file name is ifcfg-eth2.

      In the /etc/sysconfig/network directory, the third network adapter (the ifcfg-eth2 file in the preceding example) may not exist by default. Therefore, you are advised to create the file and configure the network information.
      Examples of the NIC configuration files on each network plane are as follows:
      • For the management plane:
      • For the storage plane:
      • For the iSCSI plane:
      The preceding IP addresses are examples only. Set IP addresses based on the data plan.

    5. Run the following command to edit the network configuration file:

      vi /etc/sysconfig/network/routes

      If the in-use OS does not have the routes file, the routes file will be created after you run the command. Continue with the follow-up operations.

    6. Add the default gateway information of the management plane to the configuration file.

      For example, run the following command:


    7. Save the configuration and exit the vi editor.
    8. Run the following command to restart the network service to make the configuration take effect:

      service network restart

    9. Configure the network information on all the other CVM nodes. For details, see 7 to 14.

    Installing OS Dependency Packages

    You can install the FusionStorage Agent (FSA) node only after installing the required OS dependency packages on the server.

    1. After the OS is installed, configure firewall rules for the OS by performing operations provided in FusionStorage Communication Matrix.

      Log in to the server OS using the server management IP address. Perform the following operations to disable firewalls and disable the function of enabling firewalls upon a startup. If firewalls are not disabled, communication between storage nodes fails.

      Commands for disabling firewalls are as follows:

      • OSs earlier than Red Hat Linux 6.9 or CentOS Linux 6.9:

        Stopping the firewall program: service iptables stop

        Disabling automatic firewall startup upon OS startup: chkconfig iptables off

      • Red Hat Linux 7.1~7.5, CentOS Linux 7.1~7.5, Oracle Linux or EluerOS

        Stopping the firewall program: systemctl stop firewalld

        Disabling automatic firewall startup upon OS startup: systemctl disable firewalld

      • SUSE Linux OSs:

        Stopping the firewall program:

        /etc/init.d/SuSEfirewall2_setup stop

        /etc/init.d/SuSEfirewall2_init stop

        Disabling automatic firewall startup upon OS startup:

        chkconfig SuSEfirewall2_setup off

        chkconfig SuSEfirewall2_init off

      For detailed operation guidelines, visit the official website of the OS provider.

    2. Decompress the FusionStorage Block software package.

      Obtain one of the following installation scripts for the OS dependency packages from the install_lib folder based on the OS type:

      • Red Hat Linux OSs:
      • SUSE Linux OSs:
      • CentOS Linux OSs:
      • Oracle Linux OSs:

    3. (Optional) Modify the installation script of OS dependency packages to adapt with the installation.

      Perform this step if you have already configured the software repository onsite and do not need to obtain the dependency package software by mounting the OS image file.

      Modify the installation script as follows:

      • SUSE Linux OSs: Edit the file as follows:

        • Change the REPO_NAME value to the name of the in-use software repository. You can query the software repository name by running the zypper lr command in the OS.
        • Change the ZYPPER_REPO_IS_EXIST value to 1.
      • Other OSs: Change the YUM_REPO_IS_EXIST value in the file to 1.

    4. Upload the obtained installation script to a directory, for example, /tmp, on each server.
    5. On the remote control page, mount the OS image file.

      This step is not required if you have configured the software repository onsite.

    6. Use PuTTY to log in to the first server.

      Ensure that the management IP address and username root are used to establish the connection.

      If the public and private keys are used to authenticate the login, perform the operations based on Using PuTTY to Log In to a Node in Key Pair Authentication Mode.

    7. Switch to the directory containing the installation script and run the following command to install the dependency packages:


      In this command, specifies the installation script name.

      For example, sh

      The installation process takes about 20 minutes. You can install the dependency packages for multiple servers at the same time. The time required for the installation varies depending on the hardware configurations and whether the software repository is used. If the installation is completed in a short period of time and no error message is reported, the installation is successful.


      If the server does not support the dependency package installation by mounting the image file, perform the following operations to install the dependency packages:

      1. Upload the local operating system image file to the server.
      2. Run the sh Name of the image file command to install the dependency packages.

        In the preceding command, specifies the name of the installation script.

        Example command: sh /tmp/CentOS-7-x86_64-DVD-1611.iso

      3. After the dependency packages are installed, the servers are automatically restarted.

    8. Install the dependency packages for all servers. For details, see 21 to 22.

    Set VMs to automatically start.

    1. Set the management VMs on each ESXi host to automatically start with the host.

      Management VMs include:

      • FSM VMs
      • CVMs
      • If vCenter VMs are deployed, set vCenter VMs to start with the host.

      The CVM must be ranked first among all automatic startup VMs to improve the storage service reliability.

      Figure 6-9 is an example of setting the VMs to automatically start with the host startup.

      Select Host and choose Configuration > Software > Virtual Machine Startup/Shutdown > Properties.

      In the displayed dialog box, move all the management VMs on the ESXi host to the Automatic Startup area and set Default Startup Delay to 0.

      Figure 6-9  VM automatic startup

    2. Safely restart each ESXi host to make the configuration take effect.

      Comply with the following requirements during the restart process:

      • Use the safe restart mode to restart all ESXi hosts. Otherwise, the configuration may not take effect. To safely restart the host, right-click the ESXi host on the vClient, choose Reboot.
      • If vCenter VMs are deployed on the host, restart the vCenter host at last. The vCenter host restart takes time, and you can log in to vCenter only after the vCenter host restart is complete.

      No further action is required.

Updated: 2019-06-29

Document ID: EDOC1100016637

Views: 27055

Downloads: 18

Average rating:
This Document Applies to these Products
Related Documents
Related Version
Previous Next