openstack
/
heat
Code Issues Proposed changes

Split out calculation of the next batch into a separate function 

Create a next_batch function in the scaling library that calculates only
what the next batch update should look like. For now, continue to call this
from the get_batches function so that we can verify there is no change in
behaviour. However, this will make it easier to add improvements in the
future, and to reconcile InstanceGroup with ResourceGroup.

Change-Id: I0fe25556de2a678ead14bc64f77c0466fe8f9e92
Partially-Implements: blueprint scaling-group-common
This commit is contained in:
Zane Bitter 2015-08-26 19:12:07 -04:00 committed by Steve Baker
parent 16b742c1e1
commit 661122c256
3 changed files with 188 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
heat/engine/resources/openstack/heat/instance_group.py
View File

@ -23,6 +23,7 @@ from heat.engine.resources import stack_resource
from heat.engine import rsrc_defn
from heat.engine import scheduler
from heat.scaling import lbutils
from heat.scaling import rolling_update
from heat.scaling import template
@ -337,20 +338,17 @@ class InstanceGroup(stack_resource.StackResource):
new definition in that batch (either by creating a new member or
updating an existing one).
"""
efft_bat_sz = min(batch_size, capacity)
efft_min_sz = min(min_in_service, capacity)
# effective capacity includes temporary capacity added to accommodate
# the minimum number of instances in service during update
efft_capacity = max(capacity - efft_bat_sz, efft_min_sz) + efft_bat_sz
efft_capacity = capacity
updated = 0
while updated < efft_capacity:
if updated >= efft_min_sz:
efft_capacity = capacity
efft_bat_sz = min(capacity - updated, efft_bat_sz)
yield efft_capacity, efft_bat_sz
updated += efft_bat_sz
while rolling_update.needs_update(capacity, efft_capacity, updated):
batch = rolling_update.next_batch(capacity, efft_capacity,
updated, batch_size,
min_in_service)
yield batch
efft_capacity, num_updates = batch
updated += num_updates
def _check_for_completion(self, updater):
while not self.check_update_complete(updater):

50
heat/scaling/rolling_update.py Normal file
View File

@ -0,0 +1,50 @@
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
def needs_update(targ_capacity, curr_capacity, num_up_to_date):
"""
Return whether there are more batch updates to do.
Inputs are the target size for the group, the current size of the group,
and the number of members that already have the latest definition.
"""
return not (num_up_to_date >= curr_capacity == targ_capacity)
def next_batch(targ_capacity, curr_capacity, num_up_to_date, batch_size,
min_in_service):
"""
Return details of the next batch in a batched update.
The result is a tuple containing the new size of the group and the number
of members that may receive the new definition (by a combination of
creating new members and updating existing ones).
Inputs are the target size for the group, the current size of the group,
the number of members that already have the latest definition, the batch
size, and the minimum number of members to keep in service during a rolling
update.
"""
efft_min_sz = min(min_in_service, targ_capacity)
efft_bat_sz = min(batch_size, max(targ_capacity - num_up_to_date, 0))
if num_up_to_date >= efft_min_sz:
new_capacity = targ_capacity
elif num_up_to_date == 0:
new_capacity = max(targ_capacity - efft_bat_sz,
efft_min_sz) + efft_bat_sz
else:
new_capacity = curr_capacity
return new_capacity, efft_bat_sz

128
heat/tests/autoscaling/test_rolling_update.py Normal file
View File

@ -0,0 +1,128 @@
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
from heat.tests import common
from heat.scaling import rolling_update
class TestNeedsUpdate(common.HeatTestCase):
scenarios = [
('4_4_0', dict(targ=4, curr=4, updated=0, result=True)),
('4_4_1', dict(targ=4, curr=4, updated=1, result=True)),
('4_4_3', dict(targ=4, curr=4, updated=3, result=True)),
('4_4_4', dict(targ=4, curr=4, updated=4, result=False)),
('4_4_4', dict(targ=4, curr=4, updated=5, result=False)),
('4_5_0', dict(targ=4, curr=5, updated=0, result=True)),
('4_5_1', dict(targ=4, curr=5, updated=1, result=True)),
('4_5_3', dict(targ=4, curr=5, updated=3, result=True)),
('4_5_4', dict(targ=4, curr=5, updated=4, result=True)),
('4_5_4', dict(targ=4, curr=5, updated=5, result=True)),
('4_3_0', dict(targ=4, curr=3, updated=0, result=True)),
('4_3_1', dict(targ=4, curr=3, updated=1, result=True)),
('4_3_2', dict(targ=4, curr=3, updated=2, result=True)),
('4_3_3', dict(targ=4, curr=3, updated=3, result=True)),
('4_3_4', dict(targ=4, curr=3, updated=4, result=True)),
]
def test_needs_update(self):
needs_update = rolling_update.needs_update(self.targ, self.curr,
self.updated)
self.assertEqual(self.result, needs_update)
class TestNextBatch(common.HeatTestCase):
scenarios = [
('4_4_0_1_0', dict(targ=4, curr=4, updated=0, bat_size=1, min_srv=0,
batch=(4, 1))),
('4_4_3_1_0', dict(targ=4, curr=4, updated=3, bat_size=1, min_srv=0,
batch=(4, 1))),
('4_4_0_1_4', dict(targ=4, curr=4, updated=0, bat_size=1, min_srv=4,
batch=(5, 1))),
('4_5_3_1_4', dict(targ=4, curr=5, updated=3, bat_size=1, min_srv=4,
batch=(5, 1))),
('4_5_4_1_4', dict(targ=4, curr=5, updated=4, bat_size=1, min_srv=4,
batch=(4, 0))),
('4_4_0_1_5', dict(targ=4, curr=4, updated=0, bat_size=1, min_srv=5,
batch=(5, 1))),
('4_5_3_1_5', dict(targ=4, curr=5, updated=3, bat_size=1, min_srv=5,
batch=(5, 1))),
('4_5_0_1_4', dict(targ=4, curr=5, updated=0, bat_size=1, min_srv=4,
batch=(5, 1))),
('4_5_1_1_4', dict(targ=4, curr=5, updated=1, bat_size=1, min_srv=4,
batch=(5, 1))),
('4_5_4_1_5', dict(targ=4, curr=5, updated=4, bat_size=1, min_srv=5,
batch=(4, 0))),
('4_4_0_2_0', dict(targ=4, curr=4, updated=0, bat_size=2, min_srv=0,
batch=(4, 2))),
('4_4_2_2_0', dict(targ=4, curr=4, updated=2, bat_size=2, min_srv=0,
batch=(4, 2))),
('4_4_0_2_4', dict(targ=4, curr=4, updated=0, bat_size=2, min_srv=4,
batch=(6, 2))),
('4_6_2_2_4', dict(targ=4, curr=4, updated=0, bat_size=2, min_srv=4,
batch=(6, 2))),
('4_6_4_2_4', dict(targ=4, curr=6, updated=4, bat_size=2, min_srv=4,
batch=(4, 0))),
('5_5_0_2_0', dict(targ=5, curr=5, updated=0, bat_size=2, min_srv=0,
batch=(5, 2))),
('5_5_4_2_0', dict(targ=5, curr=5, updated=4, bat_size=2, min_srv=0,
batch=(5, 1))),
('5_5_0_2_4', dict(targ=5, curr=5, updated=0, bat_size=2, min_srv=4,
batch=(6, 2))),
('5_6_2_2_4', dict(targ=5, curr=6, updated=2, bat_size=2, min_srv=4,
batch=(6, 2))),
('5_6_4_2_4', dict(targ=5, curr=6, updated=4, bat_size=2, min_srv=4,
batch=(5, 1))),
('3_3_0_2_0', dict(targ=3, curr=3, updated=0, bat_size=2, min_srv=0,
batch=(3, 2))),
('3_3_2_2_0', dict(targ=3, curr=3, updated=2, bat_size=2, min_srv=0,
batch=(3, 1))),
('3_3_0_2_4', dict(targ=3, curr=3, updated=0, bat_size=2, min_srv=4,
batch=(5, 2))),
('3_5_2_2_4', dict(targ=3, curr=5, updated=2, bat_size=2, min_srv=4,
batch=(5, 1))),
('3_5_3_2_4', dict(targ=3, curr=5, updated=3, bat_size=2, min_srv=4,
batch=(3, 0))),
('4_4_0_4_0', dict(targ=4, curr=4, updated=0, bat_size=4, min_srv=0,
batch=(4, 4))),
('4_4_0_5_0', dict(targ=4, curr=4, updated=0, bat_size=5, min_srv=0,
batch=(4, 4))),
('4_4_0_4_1', dict(targ=4, curr=4, updated=0, bat_size=4, min_srv=1,
batch=(5, 4))),
('4_4_4_4_1', dict(targ=4, curr=4, updated=4, bat_size=4, min_srv=1,
batch=(4, 0))),
('4_4_0_6_1', dict(targ=4, curr=4, updated=0, bat_size=6, min_srv=1,
batch=(5, 4))),
('4_4_4_6_1', dict(targ=4, curr=4, updated=4, bat_size=6, min_srv=1,
batch=(4, 0))),
('4_4_0_4_2', dict(targ=4, curr=4, updated=0, bat_size=4, min_srv=2,
batch=(6, 4))),
('4_4_4_4_2', dict(targ=4, curr=4, updated=4, bat_size=4, min_srv=2,
batch=(4, 0))),
('4_4_0_4_4', dict(targ=4, curr=4, updated=0, bat_size=4, min_srv=4,
batch=(8, 4))),
('4_4_4_4_4', dict(targ=4, curr=4, updated=4, bat_size=4, min_srv=4,
batch=(4, 0))),
('4_4_0_5_6', dict(targ=4, curr=4, updated=0, bat_size=5, min_srv=6,
batch=(8, 4))),
('4_4_4_5_6', dict(targ=4, curr=4, updated=4, bat_size=5, min_srv=6,
batch=(4, 0))),
]
def test_next_batch(self):
batch = rolling_update.next_batch(self.targ, self.curr,
self.updated,
self.bat_size,
self.min_srv)
self.assertEqual(self.batch, batch)