Stop using scheduler RPC API magic

If a scheduler RPC message isn't handled directly by a SchedulerManager
method, the __getattr__() fallback passes the message to a driver
method in the form of schedule_${method}() and, if that doesn't exist,
instead calls the schedule() method supplying the topic and method
args.

This is pretty bizarre stuff and we appear to only use it in two cases:

  1) live_migration - this is how the schedule_live_migration()
     method in the driver gets called, but the side-effect is that
     we require the client to pass a topic argument which is never
     used. This would be much more sanely handled with an explicit
     SchedulerManager.live_migration() method.

  2) create_volume - the volume API asks the scheduler to pick a
     target host for create_volume() using this method. This would
     be easily handled with an SchedulerManager.create_volume()
     method.

Change-Id: I1047489d85ac51d8d36fea1c4eb858df638ce349
This commit is contained in:
Mark McLoughlin
2012-08-29 09:21:03 +01:00
parent c91d2d74f0
commit ee344a3b8c
5 changed files with 28 additions and 11 deletions

View File

@@ -89,6 +89,9 @@ class MultiScheduler(driver.Scheduler):
def schedule_prep_resize(self, *args, **kwargs):
return self.drivers['compute'].schedule_prep_resize(*args, **kwargs)
def schedule_create_volume(self, *args, **kwargs):
return self.drivers['volume'].schedule_create_volume(*args, **kwargs)
def update_service_capabilities(self, service_name, host, capabilities):
# Multi scheduler is only a holder of sub-schedulers, so
# pass the capabilities to the schedulers that matter

View File

@@ -41,6 +41,7 @@ class SchedulerAPI(nova.openstack.common.rpc.proxy.RpcProxy):
1.4 - Remove update_db from prep_resize
1.5 - Add reservations argument to prep_resize()
1.6 - Remove reservations argument to run_instance()
1.7 - Add create_volume() method, remove topic from live_migration()
'''
BASE_RPC_API_VERSION = '1.0'
@@ -73,14 +74,21 @@ class SchedulerAPI(nova.openstack.common.rpc.proxy.RpcProxy):
return self.call(ctxt, self.make_msg('show_host_resources', host=host))
def live_migration(self, ctxt, block_migration, disk_over_commit,
instance, dest, topic):
instance, dest):
# NOTE(comstud): Call vs cast so we can get exceptions back, otherwise
# this call in the scheduler driver doesn't return anything.
instance_p = jsonutils.to_primitive(instance)
return self.call(ctxt, self.make_msg('live_migration',
block_migration=block_migration,
disk_over_commit=disk_over_commit, instance=instance_p,
dest=dest, topic=topic), version='1.3')
dest=dest), version='1.7')
def create_volume(self, ctxt, volume_id, snapshot_id, reservations):
self.cast(ctxt,
self.make_msg('create_volume',
volume_id=volume_id, snapshot_id=snapshot_id,
reservations=reservations),
version='1.7')
def update_service_capabilities(self, ctxt, service_name, host,
capabilities):

View File

@@ -54,7 +54,8 @@ class SimpleScheduler(chance.ChanceScheduler):
super(SimpleScheduler, self).schedule_run_instance(context,
request_spec, reservations, *_args, **_kwargs)
def schedule_create_volume(self, context, volume_id, *_args, **_kwargs):
def schedule_create_volume(self, context, volume_id, snapshot_id,
reservations):
"""Picks a host that is up and has the fewest volumes."""
deprecated.warn(_('nova-volume functionality is deprecated in Folsom '
'and will be removed in Grizzly. Volumes are now handled '
@@ -72,7 +73,8 @@ class SimpleScheduler(chance.ChanceScheduler):
if not utils.service_is_up(service):
raise exception.WillNotSchedule(host=host)
driver.cast_to_volume_host(context, host, 'create_volume',
volume_id=volume_id, **_kwargs)
volume_id=volume_id, snapshot_id=snapshot_id,
reservations=reservations)
return None
results = db.service_get_all_volume_sorted(elevated)
@@ -86,7 +88,8 @@ class SimpleScheduler(chance.ChanceScheduler):
raise exception.NoValidHost(reason=msg)
if utils.service_is_up(service) and not service['disabled']:
driver.cast_to_volume_host(context, service['host'],
'create_volume', volume_id=volume_id, **_kwargs)
'create_volume', volume_id=volume_id,
snapshot_id=snapshot_id, reservations=reservations)
return None
msg = _("Is the appropriate service running?")
raise exception.NoValidHost(reason=msg)

View File

@@ -46,9 +46,6 @@ class FakeVolumeScheduler(driver.Scheduler):
super(FakeVolumeScheduler, self).__init__()
self.is_update_caps_called = False
def schedule_create_volume(self, *args, **kwargs):
pass
def schedule_create_volumes(self, *args, **kwargs):
pass
@@ -97,7 +94,7 @@ class MultiDriverTestCase(test_scheduler.SchedulerTestCase):
#no compute methods are proxied at this time
test_methods = {compute_driver: [],
volume_driver: ['create_volume', 'create_volumes']}
volume_driver: ['create_volumes']}
for driver, methods in test_methods.iteritems():
for method in methods:

View File

@@ -86,10 +86,16 @@ class SchedulerRpcAPITestCase(test.TestCase):
self._test_scheduler_api('live_migration', rpc_method='call',
block_migration='fake_block_migration',
disk_over_commit='fake_disk_over_commit',
instance='fake_instance', dest='fake_dest', topic='fake_topic',
version='1.3')
instance='fake_instance', dest='fake_dest',
version='1.7')
def test_update_service_capabilities(self):
self._test_scheduler_api('update_service_capabilities',
rpc_method='fanout_cast', service_name='fake_name',
host='fake_host', capabilities='fake_capabilities')
def test_create_volume(self):
self._test_scheduler_api('create_volume',
rpc_method='cast', volume_id="fake_volume",
snapshot_id="fake_snapshots", reservations=list('fake_res'),
version='1.7')