Revision 57fb6fcb
| b/doc/install.rst | ||
|---|---|---|
| 421 | 421 |
|
| 422 | 422 |
**Mandatory** on all nodes. |
| 423 | 423 |
|
| 424 |
You can run Ganeti either in "bridged mode" or in "routed mode". In
|
|
| 425 |
bridged mode, the default, the instances network interfaces will be
|
|
| 426 |
attached to a software bridge running in dom0. Xen by default creates
|
|
| 427 |
such a bridge at startup, but your distribution might have a different
|
|
| 428 |
way to do things, and you'll definitely need to manually set it up under
|
|
| 429 |
KVM. |
|
| 424 |
You can run Ganeti either in "bridged mode", "routed mode" or
|
|
| 425 |
"openvswitch mode". In bridged mode, the default, the instances network
|
|
| 426 |
interfaces will be attached to a software bridge running in dom0. Xen by
|
|
| 427 |
default creates such a bridge at startup, but your distribution might
|
|
| 428 |
have a different way to do things, and you'll definitely need to
|
|
| 429 |
manually set it up under KVM.
|
|
| 430 | 430 |
|
| 431 | 431 |
Beware that the default name Ganeti uses is ``xen-br0`` (which was used |
| 432 | 432 |
in Xen 2.0) while Xen 3.0 uses ``xenbr0`` by default. See the |
| ... | ... | |
| 449 | 449 |
instances, through their interface, in the table you specified (under |
| 450 | 450 |
KVM, and in the main table under Xen). |
| 451 | 451 |
|
| 452 |
Also for "openvswitch mode" under Xen a custom network script is needed. |
|
| 453 |
Under KVM everything should work, but you'll need to configure your |
|
| 454 |
switches outside of Ganeti (as for bridges). |
|
| 455 |
|
|
| 452 | 456 |
.. admonition:: Bridging issues with certain kernels |
| 453 | 457 |
|
| 454 | 458 |
Some kernel versions (e.g. 2.6.32) have an issue where the bridge |
| b/htools/Ganeti/Types.hs | ||
|---|---|---|
| 366 | 366 |
$(THH.declareSADT "NICMode" |
| 367 | 367 |
[ ("NMBridged", 'C.nicModeBridged)
|
| 368 | 368 |
, ("NMRouted", 'C.nicModeRouted)
|
| 369 |
, ("NMOvs", 'C.nicModeOvs)
|
|
| 369 | 370 |
]) |
| 370 | 371 |
$(THH.makeJSONInstance ''NICMode) |
| 371 | 372 |
|
| b/lib/cmdlib.py | ||
|---|---|---|
| 13053 | 13053 |
raise errors.OpPrereqError("Cannot set the NIC IP address to None"
|
| 13054 | 13054 |
" on a routed NIC", errors.ECODE_INVAL) |
| 13055 | 13055 |
|
| 13056 |
elif new_mode == constants.NIC_MODE_OVS: |
|
| 13057 |
# TODO: check OVS link |
|
| 13058 |
self.LogInfo("OVS links are currently not checked for correctness")
|
|
| 13059 |
|
|
| 13056 | 13060 |
if constants.INIC_MAC in params: |
| 13057 | 13061 |
mac = params[constants.INIC_MAC] |
| 13058 | 13062 |
if mac is None: |
| b/lib/constants.py | ||
|---|---|---|
| 1068 | 1068 |
|
| 1069 | 1069 |
NIC_MODE_BRIDGED = "bridged" |
| 1070 | 1070 |
NIC_MODE_ROUTED = "routed" |
| 1071 |
NIC_MODE_OVS = "openvswitch" |
|
| 1071 | 1072 |
NIC_IP_POOL = "pool" |
| 1072 | 1073 |
|
| 1073 |
NIC_VALID_MODES = frozenset([NIC_MODE_BRIDGED, NIC_MODE_ROUTED]) |
|
| 1074 |
NIC_VALID_MODES = frozenset([NIC_MODE_BRIDGED, NIC_MODE_ROUTED, NIC_MODE_OVS])
|
|
| 1074 | 1075 |
|
| 1075 | 1076 |
RESERVE_ACTION = "reserve" |
| 1076 | 1077 |
RELEASE_ACTION = "release" |
| b/man/ganeti-os-interface.rst | ||
|---|---|---|
| 101 | 101 |
instance. |
| 102 | 102 |
|
| 103 | 103 |
NIC_%N_MODE |
| 104 |
The NIC mode, either routed or bridged
|
|
| 104 |
The NIC mode, routed, bridged or openvswitch
|
|
| 105 | 105 |
|
| 106 | 106 |
NIC_%N_BRIDGE |
| 107 | 107 |
The bridge to which this NIC will be attached. This variable is |
| 108 | 108 |
defined only when the NIC is in bridged mode. |
| 109 | 109 |
|
| 110 | 110 |
NIC_%N_LINK |
| 111 |
If the NIC is in bridged mode, this is the same as
|
|
| 112 |
``NIC_%N_BRIDGE``. If it is in routed mode, the routing table
|
|
| 113 |
which will be used by the hypervisor to insert the appropriate
|
|
| 114 |
routes. |
|
| 111 |
In bridged or openvswitch mode, this is the interface to which the
|
|
| 112 |
NIC will be attached (same as ``NIC_%N_BRIDGE`` for bridged). In
|
|
| 113 |
routed mode it is the routing table which will be used by the
|
|
| 114 |
hypervisor to insert the appropriate routes.
|
|
| 115 | 115 |
|
| 116 | 116 |
NIC_%N_FRONTEND_TYPE |
| 117 | 117 |
(Optional) If applicable, the type of the exported NIC to the |
| b/man/gnt-backup.rst | ||
|---|---|---|
| 121 | 121 |
the node expects the instance to use) |
| 122 | 122 |
|
| 123 | 123 |
mode |
| 124 |
specifies the connection mode for this nic: ``routed`` or
|
|
| 125 |
``bridged``.
|
|
| 124 |
specifies the connection mode for this nic: ``routed``,
|
|
| 125 |
``bridged`` or ``openvswitch``
|
|
| 126 | 126 |
|
| 127 | 127 |
link |
| 128 |
in bridged mode specifies the bridge to attach this NIC to, in
|
|
| 129 |
routed mode it's intended to differentiate between different
|
|
| 130 |
routing tables/instance groups (but the meaning is dependent on
|
|
| 131 |
the network script in use, see **gnt-cluster**(8) for more
|
|
| 132 |
details) |
|
| 128 |
in bridged and openvswitch mode specifies the interface to attach
|
|
| 129 |
this NIC to, in routed mode it's intended to differentiate between
|
|
| 130 |
different routing tables/instance groups (but the meaning is
|
|
| 131 |
dependent on the network script in use, see **gnt-cluster**(8) for
|
|
| 132 |
more details)
|
|
| 133 | 133 |
|
| 134 | 134 |
Of these ``mode`` and ``link`` are nic parameters, and inherit their |
| 135 | 135 |
default at cluster level. |
| b/man/gnt-cluster.rst | ||
|---|---|---|
| 363 | 363 |
list of key=value pairs with the following supported keys: |
| 364 | 364 |
|
| 365 | 365 |
mode |
| 366 |
The default nic mode, 'routed' or 'bridged'.
|
|
| 366 |
The default nic mode, 'routed', 'bridged' or 'openvswitch'.
|
|
| 367 | 367 |
|
| 368 | 368 |
link |
| 369 |
In bridged mode the default NIC bridge. In routed mode it |
|
| 370 |
represents an hypervisor-vif-script dependent value to allow |
|
| 371 |
different instance groups. For example under the KVM default |
|
| 372 |
network script it is interpreted as a routing table number or |
|
| 373 |
name. |
|
| 369 |
In bridged or openvswitch mode the default interface where to attach |
|
| 370 |
NICs. In routed mode it represents an hypervisor-vif-script |
|
| 371 |
dependent value to allow different instance groups. For example |
|
| 372 |
under the KVM default network script it is interpreted as a routing |
|
| 373 |
table number or name. Openvswitch support is also hypervisor |
|
| 374 |
dependent and currently works for the default KVM network script. |
|
| 375 |
Under Xen a custom network script must be provided. |
|
| 374 | 376 |
|
| 375 | 377 |
The ``-D (--disk-parameters)`` option allows you to set the default disk |
| 376 | 378 |
template parameters at cluster level. The format used for this option is |
| b/man/gnt-instance.rst | ||
|---|---|---|
| 103 | 103 |
the node expects the instance to use) |
| 104 | 104 |
|
| 105 | 105 |
mode |
| 106 |
specifies the connection mode for this nic: routed or bridged. |
|
| 106 |
specifies the connection mode for this nic: routed, bridged or |
|
| 107 |
openvswitch. |
|
| 107 | 108 |
|
| 108 | 109 |
link |
| 109 |
in bridged mode specifies the bridge to attach this NIC to, in |
|
| 110 |
routed mode it's intended to differentiate between different |
|
| 111 |
routing tables/instance groups (but the meaning is dependent on |
|
| 112 |
the network script, see gnt-cluster(8) for more details) |
|
| 110 |
in bridged or openvswitch mode specifies the interface to attach |
|
| 111 |
this NIC to, in routed mode it's intended to differentiate between |
|
| 112 |
different routing tables/instance groups (but the meaning is |
|
| 113 |
dependent on the network script, see gnt-cluster(8) for more |
|
| 114 |
details). Note that openvswitch support is also hypervisor |
|
| 115 |
dependent. |
|
| 113 | 116 |
|
| 114 | 117 |
|
| 115 | 118 |
Of these "mode" and "link" are nic parameters, and inherit their |
| b/tools/kvm-ifup.in | ||
|---|---|---|
| 47 | 47 |
|
| 48 | 48 |
# Connect the interface to the bridge |
| 49 | 49 |
brctl addif $BRIDGE $INTERFACE |
| 50 |
|
|
| 51 |
elif [ "$MODE" = "openvswitch" ]; then |
|
| 52 |
ovs-vsctl add-port ${LINK} $INTERFACE
|
|
| 53 |
|
|
| 50 | 54 |
else |
| 51 | 55 |
ip link set $INTERFACE up |
| 52 | 56 |
|
| ... | ... | |
| 65 | 69 |
ip route replace $IP proto static dev $INTERFACE |
| 66 | 70 |
fi |
| 67 | 71 |
|
| 72 |
# Allow routing and arp proxying, or ndp proxying (IPv6) |
|
| 68 | 73 |
if [ -d "/proc/sys/net/ipv4/conf/$INTERFACE" ]; then |
| 69 | 74 |
echo 1 > /proc/sys/net/ipv4/conf/$INTERFACE/proxy_arp |
| 70 | 75 |
echo 1 > /proc/sys/net/ipv4/conf/$INTERFACE/forwarding |
Also available in: Unified diff