Add Juju Network Space support

Juju 2.0 provides support for network spaces, allowing
charm authors to support direct binding of relations and
extra-bindings onto underlying network spaces.

Resync charm-helpers to pickup support in API endpoint
resolution code and add API extra-bindings to the charm
metadata.

Change-Id: Ibbbeba5966f799d25d0a40b69b1672deb1476924
This commit is contained in:
James Page 2016-03-31 08:45:24 +01:00
parent f434b40b04
commit ffd4644a10
6 changed files with 79 additions and 9 deletions

View File

@ -212,3 +212,28 @@ The following is a full list of current tip repos (may not be up-to-date):
- {name: cinder,
repository: 'git://github.com/openstack/cinder',
branch: master}
Network Space support
---------------------
This charm supports the use of Juju Network Spaces, allowing the charm to be bound to network space configurations managed directly by Juju. This is only supported with Juju 2.0 and above.
API endpoints can be bound to distinct network spaces supporting the network separation of public, internal and admin endpoints.
To use this feature, use the --bind option when deploying the charm:
juju deploy cinder --bind "public=public-space internal=internal-space admin=admin-space"
alternatively these can also be provided as part of a juju native bundle configuration:
cinder:
charm: cs:xenial/cinder
num_units: 1
bindings:
public: public-space
admin: admin-space
internal: internal-space
NOTE: Spaces must be configured in the underlying provider prior to attempting to use them.
NOTE: Existing deployments using os-*-network configuration options will continue to function; these options are preferred over any network space binding provided if set.

View File

@ -191,6 +191,15 @@ get_iface_for_address = partial(_get_for_address, key='iface')
get_netmask_for_address = partial(_get_for_address, key='netmask')
def resolve_network_cidr(ip_address):
'''
Resolves the full address cidr of an ip_address based on
configured network interfaces
'''
netmask = get_netmask_for_address(ip_address)
return str(netaddr.IPNetwork("%s/%s" % (ip_address, netmask)).cidr)
def format_ipv6_addr(address):
"""If address is IPv6, wrap it in '[]' otherwise return None.

View File

@ -126,7 +126,9 @@ class OpenStackAmuletDeployment(AmuletDeployment):
# Charms which can not use openstack-origin, ie. many subordinates
no_origin = ['cinder-ceph', 'hacluster', 'neutron-openvswitch', 'nrpe',
'openvswitch-odl', 'neutron-api-odl', 'odl-controller',
'cinder-backup']
'cinder-backup', 'nexentaedge-data',
'nexentaedge-iscsi-gw', 'nexentaedge-swift-gw',
'cinder-nexentaedge', 'nexentaedge-mgmt']
if self.openstack:
for svc in services:

View File

@ -14,16 +14,19 @@
# You should have received a copy of the GNU Lesser General Public License
# along with charm-helpers. If not, see <http://www.gnu.org/licenses/>.
from charmhelpers.core.hookenv import (
config,
unit_get,
service_name,
network_get_primary_address,
)
from charmhelpers.contrib.network.ip import (
get_address_in_network,
is_address_in_network,
is_ipv6,
get_ipv6_addr,
resolve_network_cidr,
)
from charmhelpers.contrib.hahelpers.cluster import is_clustered
@ -33,16 +36,19 @@ ADMIN = 'admin'
ADDRESS_MAP = {
PUBLIC: {
'binding': 'public',
'config': 'os-public-network',
'fallback': 'public-address',
'override': 'os-public-hostname',
},
INTERNAL: {
'binding': 'internal',
'config': 'os-internal-network',
'fallback': 'private-address',
'override': 'os-internal-hostname',
},
ADMIN: {
'binding': 'admin',
'config': 'os-admin-network',
'fallback': 'private-address',
'override': 'os-admin-hostname',
@ -110,7 +116,7 @@ def resolve_address(endpoint_type=PUBLIC):
correct network. If clustered with no nets defined, return primary vip.
If not clustered, return unit address ensuring address is on configured net
split if one is configured.
split if one is configured, or a Juju 2.0 extra-binding has been used.
:param endpoint_type: Network endpoing type
"""
@ -125,23 +131,45 @@ def resolve_address(endpoint_type=PUBLIC):
net_type = ADDRESS_MAP[endpoint_type]['config']
net_addr = config(net_type)
net_fallback = ADDRESS_MAP[endpoint_type]['fallback']
binding = ADDRESS_MAP[endpoint_type]['binding']
clustered = is_clustered()
if clustered:
if not net_addr:
# If no net-splits defined, we expect a single vip
resolved_address = vips[0]
else:
if clustered and vips:
if net_addr:
for vip in vips:
if is_address_in_network(net_addr, vip):
resolved_address = vip
break
else:
# NOTE: endeavour to check vips against network space
# bindings
try:
bound_cidr = resolve_network_cidr(
network_get_primary_address(binding)
)
for vip in vips:
if is_address_in_network(bound_cidr, vip):
resolved_address = vip
break
except NotImplementedError:
# If no net-splits configured and no support for extra
# bindings/network spaces so we expect a single vip
resolved_address = vips[0]
else:
if config('prefer-ipv6'):
fallback_addr = get_ipv6_addr(exc_list=vips)[0]
else:
fallback_addr = unit_get(net_fallback)
resolved_address = get_address_in_network(net_addr, fallback_addr)
if net_addr:
resolved_address = get_address_in_network(net_addr, fallback_addr)
else:
# NOTE: only try to use extra bindings if legacy network
# configuration is not in use
try:
resolved_address = network_get_primary_address(binding)
except NotImplementedError:
resolved_address = fallback_addr
if resolved_address is None:
raise ValueError("Unable to resolve a suitable IP address based on "

View File

@ -7,6 +7,10 @@ tags:
- openstack
- storage
- misc
extra-bindings:
public:
admin:
internal:
provides:
nrpe-external-master:
interface: nrpe-external-master

View File

@ -126,7 +126,9 @@ class OpenStackAmuletDeployment(AmuletDeployment):
# Charms which can not use openstack-origin, ie. many subordinates
no_origin = ['cinder-ceph', 'hacluster', 'neutron-openvswitch', 'nrpe',
'openvswitch-odl', 'neutron-api-odl', 'odl-controller',
'cinder-backup']
'cinder-backup', 'nexentaedge-data',
'nexentaedge-iscsi-gw', 'nexentaedge-swift-gw',
'cinder-nexentaedge', 'nexentaedge-mgmt']
if self.openstack:
for svc in services: