Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: NIOSSPT-16398: Updated the limitations that need writethrough to be enabled.

vNIOS for KVM supports most of the features of the Infoblox NIOS appliances, with the following limitations:

...

  • If the vNIOS instance fails to launch, the ports created for the instance are not deleted automatically. You need to delete them manually.
  • When you start an FTP connection on OpenStack with the Listen on port field set to 2021, you need to manually add a rule to allow the 2021 port. This ensures that the connection is successful.
  • Connection to the FTP service might fail when the virtual appliance enter the passive mode. To avoid this, do the following:
    • Use active mode instead of passive mode.
    • Modify the vnios-sec-group security group to open ports 1023 and above.
    • Use the FTP client inside the internal network.
  • The vNIOS instances deployed in the OpenStack environment do not support HSM Safenet groups.
  • IPv6 support in the OpenStack environment is limited by the Juno release and has not been verified by Infoblox.
  • You cannot configure the reporting virtual appliance as an HA pair. You also cannot configure it as a Grid Master or Grid Master Candidate. You can use it only as a dedicated reporting server in the Grid.
  • At certain times, a power  power failure (or an unexpected reboot) of the VM host may cause a corruption of the VM guest corruption unless writes are correctly persisted.   To do this, you must have an NVRAM-based disk controller on the host system and turn-off disk caching for the guest on the VM host. That is, you must set the value of the disk_cachemodes parameter to the  writethrough value. The  The method to set the the disk_cachemodes parameter varies for different hypervisors. 
    • The following example pertains to vNIOS
    for KVM
    • for KVM deployed in OpenStack 16.2
    • :
        Log
        1. Sign in to the system as a root user.

        2. Edit

        the edit
        1. the nova.conf

        and nova-cpu.conf
        1. files by running

        the vim
        1. the following commands:
          vi /var/lib/config-data/puppet-generated/nova/etc/nova/nova.conf

        and vim

        1. vi /var/lib/config-data/puppet-generated/nova_libvirt/etc/nova/nova

        -cpu
        1. .conf

        commands.
      1. Go to the [libvirt] section in both the files.
        1. On a new line,

        add 
        1. add:
          disk_cachemodes = file=

        "writethrough".Restart the nova services.
        1. writethrough,block=writeback,network=writeback

        2. Save the changes and restart all containers related to Nova using the command:
          podman restart container_id

      2. The following example pertains to vNIOS
    for KVM Hypervisor through the KVM Manager:
    • In KVM virt-manager, select writethrough from the disk cachemode drop-down list when deploying the instance.
    • Validate the instance by editing the guest XML file and updating for KVM Hypervisor:
      1. Set the cache mode to writethrough by passing the cache=writethrough parameter in the virt-install command:
      2. Validate this setting in the instance by editing the guest XML file as follows:
        Configure the cache mode by updating the cache parameter inside the driver
       tag
      1.  tag to specify a caching
      option to configure the cache mode.
      1. option.
        For example, to set
      the
      1. the cache
      parameter to
      1.  parameter to writethrough
        <disk type='file' device='disk'>
        <driver name='qemu' type='qcow2' cache='writethrough'/>
        <source file='/home/vms/mygm.qcow2' index='1'/>
        <backingStore/>
        <target dev='vda' bus='virtio'/>
        <alias name='virtio-disk0'/>
        ><address type='pci' domain='0x0000' bus='0x07' slot='0x00' function='0x0'/>
        </disk>