117 lines
6.4 KiB
YAML
117 lines
6.4 KiB
YAML
attributes:
|
|
sriov:
|
|
value: false
|
|
label: "SR-IOV direct port creation in private VLAN networks (Neutron)"
|
|
weight: 11
|
|
type: "checkbox"
|
|
description: |
|
|
If selected, a Neutron SR-IOV driver will be configured to enable SR-IOV capabilities using Neutron direct port.
|
|
This feature is only supported if the intel_iommu=on parameter is appended to the kernel parameters.
|
|
In case the NEO SDN driver is not used for private network VLAN provisioning:
|
|
* When in Ethernet mode, the switch side ports must be pre-configured with the required VLAN range
|
|
* When in InfinBand mode, the Subnet Manager partitions.conf file must include the private network conversions of VLAN to PKey
|
|
restrictions:
|
|
- condition: "settings:common.libvirt_type.value != 'kvm' or not ('intel_iommu=on' in settings:kernel_params.kernel.value) or not (cluster:net_provider == 'neutron' and networking_parameters:segmentation_type == 'vlan')"
|
|
message: "For Mellanox SR-IOV support in Neutron, the hypervisor type should be KVM, the environment networking should be based on Neutron with VLAN segmentation and the intel_iommu=on parameter should be appended to the kernel parameters."
|
|
num_of_vfs:
|
|
value: "16"
|
|
label: "Number of virtual NICs"
|
|
description: "One virtual function might be reserved for the storage network, if choosing iSER"
|
|
weight: 12
|
|
type: "text"
|
|
regex:
|
|
source: "^(6[0-2]|[1-5][0-9]|[1-9])$"
|
|
error: "Invalid Virtual NICs number. Please use a number in the range of 1-62."
|
|
restrictions:
|
|
- condition: "settings:mellanox-plugin.sriov.value == false"
|
|
message: "In order to change the number of virtual NICs, Neutron SR-IOV plugin should be checked."
|
|
|
|
reboot_required:
|
|
value: false
|
|
label: "Reboot when maximum number of vfs configured by the user is greater than the one burned in FW."
|
|
description: |
|
|
Reboots Slave nodes if the maximum number of vfs configured by the user is greator than the number burned in FW.
|
|
weight: 30
|
|
type: hidden
|
|
|
|
mlnx_qos:
|
|
value: false
|
|
label: "Quality of Service over Mellanox SR-IOV ports (Neutron)"
|
|
description: |
|
|
If selected, Neutron "Quality of Service" (QoS) will be enabled for SR-IOV ports over Mellanox HCAs.
|
|
This feature is supported in Ethernet mode over Neutron with VLAN after SR-IOV direct port creation.
|
|
weight: 30
|
|
type: hidden
|
|
restrictions:
|
|
- condition: "settings:common.libvirt_type.value != 'kvm' or not (settings:mellanox-plugin.sriov.value == true) or not (cluster:net_provider == 'neutron' and networking_parameters:segmentation_type == 'vlan')"
|
|
message: "For Mellanox QoS support in Neutron, hypervisor type should be KVM and the environment networking should be based on Neutron with VLAN segmentation with Mellanox SR-IOV support."
|
|
|
|
use_mlnx_neo:
|
|
value: false
|
|
label: "NEO SDN controller auto VLAN Provisioning (Neutron)"
|
|
weight: 40
|
|
type: "checkbox"
|
|
description: |
|
|
If selected, Mellanox NEO driver will be used to support auto-switch VLAN or SM Pkeys provisioning.
|
|
This feature is supported over Neutron with VLAN.
|
|
restrictions:
|
|
- condition: "settings:common.libvirt_type.value != 'kvm' or not (cluster:net_provider == 'neutron' and networking_parameters:segmentation_type == 'vlan')"
|
|
message: "For NEO SDN auto VLAN provisioning, hypervisor type should be KVM and the environment networking should be based on Neutron with VLAN segmentation."
|
|
mlnx_neo_ip:
|
|
value: ""
|
|
label: "NEO IP"
|
|
weight: 41
|
|
type: "text"
|
|
regex:
|
|
source: '^(((25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.){3}(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?))$'
|
|
error: "Invalid IP address"
|
|
restrictions:
|
|
- condition: "settings:mellanox-plugin.use_mlnx_neo.value == false"
|
|
message: "In order to insert NEO SDN Controller IP, Mellanox NEO auto VLAN provisioning has to be checked."
|
|
mlnx_neo_user:
|
|
value: "admin"
|
|
label: "NEO username"
|
|
weight: 42
|
|
type: "text"
|
|
restrictions:
|
|
- condition: "settings:mellanox-plugin.use_mlnx_neo.value == false"
|
|
message: "In order to insert NEO SDN Controller username, Mellanox NEO auto VLAN provisioning has to be checked."
|
|
mlnx_neo_password:
|
|
value: "123456"
|
|
label: "NEO password"
|
|
weight: 43
|
|
type: "text"
|
|
restrictions:
|
|
- condition: "settings:mellanox-plugin.use_mlnx_neo.value == false"
|
|
message: "In order to insert NEO SDN Controller password, Mellanox NEO auto VLAN provisioning has to be checked."
|
|
|
|
vxlan_offloading:
|
|
value: false
|
|
label: "VXLAN Offloading (Neutron)"
|
|
description: |
|
|
If selected, Mellanox hardware will be used to achieve a better performance and significant CPU overhead reduction
|
|
using VXLAN traffic offloading. This feature is supported in Ethernet mode over Neutron with tunneling.
|
|
weight: 50
|
|
type: "checkbox"
|
|
restrictions:
|
|
- condition: "not (cluster:net_provider == 'neutron' and networking_parameters:segmentation_type == 'tun')"
|
|
message: "For VXLAN offloading, network should be neutron with tunnling segmentation."
|
|
|
|
iser:
|
|
value: false
|
|
label: "iSER protocol for volumes (Cinder)"
|
|
description: |
|
|
High Performance Block Storage: Cinder volumes over iSER protocol (iSCSI over RDMA).
|
|
In Ethernet mode, a dedicated Virtual Function will be reserved for a storage endpoint, and the priority flow control
|
|
has to be enabled on the switch side port.
|
|
In InfiniBand mode, the IPoIB parent interface of the network storage interface will be used as the storage endpoint.
|
|
|
|
|
|
For more information about Mellanox NEO installation and usage: https://community.mellanox.com/docs/DOC-2155
|
|
For more information about Mellanox Fuel 8.0 plugin configurations: https://community.mellanox.com/docs/DOC-2435
|
|
weight: 60
|
|
type: "checkbox"
|
|
restrictions:
|
|
- condition: "settings:storage.volumes_lvm.value != true or settings:common.libvirt_type.value != 'kvm' or not ('intel_iommu=on' in settings:kernel_params.kernel.value)"
|
|
message: "For iSER support in Cinder, Cinder LVM over iSCSI should be enabled, the hypervisor type should be KVM and the intel_iommu=on parameter should be appended to the kernel parameters."
|