forked from openstack-archive/fuel-plugin-mellanox
-
Notifications
You must be signed in to change notification settings - Fork 0
/
environment_config.yaml
116 lines (109 loc) · 6.38 KB
/
environment_config.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
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: hidden
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."