def __init__(self, application_id, plugin_info):
self.logger = ScalingLog("proportional_derivative.controller.log",
"controller.log", application_id)
plugin_info = plugin_info["plugin_info"]
self.application_id = application_id
self.instances = plugin_info["instances"]
self.check_interval = plugin_info["check_interval"]
self.trigger_down = plugin_info["trigger_down"]
self.trigger_up = plugin_info["trigger_up"]
self.min_cap = plugin_info["min_cap"]
self.max_cap = plugin_info["max_cap"]
self.metric_rounding = plugin_info["metric_rounding"]
self.actuator_type = plugin_info["actuator"]
self.metric_source_type = plugin_info["metric_source"]
self.heuristic_options = plugin_info["heuristic_options"]
self.running = True
self.running_lock = threading.RLock()
# Gets a new metric source plugin using the given name
metric_source = MetricSourceBuilder().get_metric_source(
self.metric_source_type, plugin_info)
# Gets a new actuator plugin using the given name
actuator = ActuatorBuilder().get_actuator(self.actuator_type,
plugin_info)
""" The alarm here is responsible for deciding whether to
scale up or down, or even do nothing """
self.alarm = ProportionalDerivativeAlarm(
actuator, metric_source, self.trigger_down, self.trigger_up,
self.min_cap, self.max_cap, self.metric_rounding,
self.heuristic_options, application_id, self.instances)
class ProportionalDerivativeController(Controller):
def __init__(self, application_id, plugin_info):
self.logger = ScalingLog("proportional_derivative.controller.log",
"controller.log", application_id)
plugin_info = plugin_info["plugin_info"]
self.application_id = application_id
self.instances = plugin_info["instances"]
self.check_interval = plugin_info["check_interval"]
self.trigger_down = plugin_info["trigger_down"]
self.trigger_up = plugin_info["trigger_up"]
self.min_cap = plugin_info["min_cap"]
self.max_cap = plugin_info["max_cap"]
self.metric_rounding = plugin_info["metric_rounding"]
self.actuator_type = plugin_info["actuator"]
self.metric_source_type = plugin_info["metric_source"]
self.heuristic_options = plugin_info["heuristic_options"]
self.running = True
self.running_lock = threading.RLock()
# Gets a new metric source plugin using the given name
metric_source = MetricSourceBuilder().get_metric_source(
self.metric_source_type, plugin_info)
# Gets a new actuator plugin using the given name
actuator = ActuatorBuilder().get_actuator(self.actuator_type,
plugin_info)
""" The alarm here is responsible for deciding whether to
scale up or down, or even do nothing """
self.alarm = ProportionalDerivativeAlarm(
actuator, metric_source, self.trigger_down, self.trigger_up,
self.min_cap, self.max_cap, self.metric_rounding,
self.heuristic_options, application_id, self.instances)
def start_application_scaling(self):
run = True
while run:
self.logger.log("Monitoring application: %s" %
(self.application_id))
# Call the alarm to check the application
try:
self.alarm.check_application_state()
except MetricNotFoundException:
self.logger.log("No metrics available")
except Exception as e:
self.logger.log(str(e))
# Wait some time
time.sleep(float(self.check_interval))
with self.running_lock:
run = self.running
def stop_application_scaling(self):
with self.running_lock:
self.running = False
def status(self):
return self.alarm.status()
def test_alarm_smoothly_negative_to_positive_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = 7.5
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, 0, 0, self.min_cap,
self.max_cap, self.metric_round, self.heuristic_options,
self.application_id_7, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 7.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = 1.5
# Derivative effect = -2.0
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 0.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -3.0
# Derivative effect = -1.5
#
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 4.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
def test_alarm_scale_up_limits(self):
#
# First call - there is no derivative effect - timestamp_1
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_5, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_5})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.max_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_5})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.max_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
def test_alarm_gets_metrics_and_scales_down_increasing_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = -45
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_3, self.instances)
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = -60
# Derivative effect = -5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 65
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -82.5
# Derivative effect = -7.5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 90
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
class TestProportionalDerivativeAlarm(unittest.TestCase):
def setUp(self):
self.application_id_0 = "app-00"
self.application_id_1 = "app-01"
self.application_id_2 = "app-02"
self.application_id_3 = "app-03"
self.application_id_4 = "app-04"
self.application_id_5 = "app-05"
self.application_id_6 = "app-06"
self.application_id_7 = "app-07"
self.timestamp_1 = datetime.datetime.strptime("2000-01-01T00:00:00.0Z",
"%Y-%m-%dT%H:%M:%S.%fZ")
self.timestamp_2 = datetime.datetime.strptime("2000-01-01T00:05:00.0Z",
"%Y-%m-%dT%H:%M:%S.%fZ")
self.timestamp_3 = datetime.datetime.strptime("2000-01-01T00:10:00.0Z",
"%Y-%m-%dT%H:%M:%S.%fZ")
self.timestamp_4 = datetime.datetime.strptime("2000-01-01T00:15:00.0Z",
"%Y-%m-%dT%H:%M:%S.%fZ")
self.timestamps = [
self.timestamp_1, self.timestamp_2, self.timestamp_3,
self.timestamp_4
]
self.instance_name_1 = "instance1"
self.instance_name_2 = "instance2"
self.instances = [self.instance_name_1, self.instance_name_2]
self.trigger_down = 10.0
self.trigger_up = 10.0
self.min_cap = 10.0
self.max_cap = 100.0
self.actuation_size = 10.0
self.allocated_resources_scale_down = 100
self.allocated_resources_scale_up = 10
self.metric_round = 2
self.default_io_cap = 34
self.bigsea_username = "******"
self.bigsea_password = "******"
# self.authorization_url = "authorization_url"
# self.authorization_data = dict(authorization_url=self.authorization_url,
# bigsea_username=self.bigsea_username,
# bigsea_password=self.bigsea_password)
compute_nodes = []
compute_nodes_key = "key"
self.instances = [self.instance_name_1, self.instance_name_2]
self.metric_source = MetricSourceBuilder().get_metric_source("nop", {})
self.instance_locator = InstanceLocator(SSHUtils({}), compute_nodes,
compute_nodes_key)
self.remote_kvm = RemoteKVM(SSHUtils({}), compute_nodes_key)
self.actuator = KVMActuator(
self.instance_locator,
self.remote_kvm, # self.authorization_data,
self.default_io_cap)
self.proportional_factor = 1.5
self.derivative_factor = 0.5
self.heuristic_options = {
"heuristic_name": "error_proportional_derivative",
"proportional_factor": self.proportional_factor,
"derivative_factor": self.derivative_factor
}
self.progress_error = {
# CASE 1
self.application_id_0: {
self.timestamp_1: -30.0,
self.timestamp_2: -20.0,
self.timestamp_3: -15.0
},
# CASE 2
self.application_id_1: {
self.timestamp_1: -30.0,
self.timestamp_2: -40.0,
self.timestamp_3: -60.0
},
# CASE 3
self.application_id_2: {
self.timestamp_1: 30.0,
self.timestamp_2: 20.0,
self.timestamp_3: 15.0
},
# CASE 4
self.application_id_3: {
self.timestamp_1: 30.0,
self.timestamp_2: 40.0,
self.timestamp_3: 55.0
},
# CASE 5
self.application_id_4: {
self.timestamp_1: 100.0,
self.timestamp_2: 100.0,
self.timestamp_3: 100.0
},
# CASE 6
self.application_id_5: {
self.timestamp_1: -100.0,
self.timestamp_2: -100.0,
self.timestamp_3: -100.0
},
# CASE 7
self.application_id_6: {
self.timestamp_1: -30.0,
self.timestamp_2: 10.0,
self.timestamp_3: 30.0
},
# CASE 8
self.application_id_7: {
self.timestamp_1: -5.0,
self.timestamp_2: -1.0,
self.timestamp_3: 2.0
}
}
def metrics(self, metric_name, options):
application_id = options["application_id"]
timestamp_to_use = self.timestamps.pop(0)
if metric_name == ProportionalDerivativeAlarm.ERROR_METRIC_NAME:
return timestamp_to_use, self.progress_error[application_id][
timestamp_to_use]
'''
CASE 1
The error is always negative and its absolute value decreases throughout the execution
The derivative component decreases the proportional effect
Uses application_id_0 progress
'''
def test_alarm_gets_metrics_and_scales_up_decreasing_error(self):
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_0, self.instances)
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = 45
# Derivative effect = 0
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_0})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = 30
# Derivative effect = -5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_0})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 25
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = 22.5
# Derivative effect = -2.5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_0})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 20
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 2
The error is always negative and its absolute value increases throughout the execution
The derivative component increases the proportional effect
Uses application_id_1 progress
'''
def test_alarm_gets_metrics_and_scales_up_increasing_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = 45
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_1, self.instances)
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_1})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = 60
# Derivative effect = 5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_1})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 65
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = 90
# Derivative effect = 10
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_1})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 90
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 3
The error is always positive and its absolute value decreases throughout the execution
The derivative component decreases the proportional effect
Uses application_id_2 progress
'''
def test_alarm_gets_metrics_and_scales_down_decreasing_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = -45
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_2, self.instances)
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_2})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = -30
# Derivative effect = 5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_2})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 25
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -22.5
# Derivative effect = 2.5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_2})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 20
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 4
The error is always positive and its absolute value increases throughout the execution
The derivative component increases the proportional effect
Uses application_id_3 progress
'''
def test_alarm_gets_metrics_and_scales_down_increasing_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = -45
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_3, self.instances)
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = -60
# Derivative effect = -5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 65
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -82.5
# Derivative effect = -7.5
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_3})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 90
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 5
The error value is set to force the use of the minimum possible cap value
Uses application_id_4 progress
'''
def test_alarm_scale_down_limits(self):
#
# First call - there is no derivative effect - timestamp_1
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_4, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_4})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.min_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_4})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.min_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 6
The error value is set to force the use of the maximum possible cap value
Uses application_id_5 progress
'''
def test_alarm_scale_up_limits(self):
#
# First call - there is no derivative effect - timestamp_1
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, self.trigger_down,
self.trigger_up, self.min_cap, self.max_cap, self.metric_round,
self.heuristic_options, self.application_id_5, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_5})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.max_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_5})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.max_cap
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 7
The error is negative in the first call and positive in the following calls.
Uses application_id_6 progress
'''
def test_alarm_negative_to_positive_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = 45
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, 0, 0, self.min_cap,
self.max_cap, self.metric_round, self.heuristic_options,
self.application_id_6, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_6})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 45
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = -15
# Derivative effect = -20
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_6})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 35
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -45
# Derivative effect = -10
#
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_6})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 55
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
'''
CASE 8
The error is negative in the two first calls and positive in the following calls.
Uses application_id_7 progress
'''
def test_alarm_smoothly_negative_to_positive_error(self):
#
# First call - there is no derivative effect - timestamp_1
# Proportional effect = 7.5
# Derivative effect = 0
#
self.alarm = ProportionalDerivativeAlarm(
self.actuator, self.metric_source, 0, 0, self.min_cap,
self.max_cap, self.metric_round, self.heuristic_options,
self.application_id_7, self.instances)
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_up)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_up + 7.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Second call - timestamp_2
# Proportional effect = 1.5
# Derivative effect = -2.0
#
# Set up mocks
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 0.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})
#
# Third call - timestamp_3
# Proportional effect = -3.0
# Derivative effect = -1.5
#
self.metric_source.get_most_recent_value = MagicMock()
self.metric_source.get_most_recent_value.side_effect = self.metrics
self.actuator.adjust_resources = MagicMock(return_value=None)
self.actuator.get_allocated_resources_to_cluster = \
MagicMock(return_value=self.allocated_resources_scale_down)
self.alarm.check_application_state()
self.metric_source.get_most_recent_value.\
assert_any_call(ProportionalDerivativeAlarm.ERROR_METRIC_NAME,
{"application_id": self.application_id_7})
# The method tries to get the amount of allocated resources
self.actuator.get_allocated_resources_to_cluster.assert_called_once_with(
self.instances)
new_cap = self.allocated_resources_scale_down - 4.5
# The method tries to adjust the amount of resources
self.actuator.adjust_resources.assert_called_once_with({
self.instance_name_1:
new_cap,
self.instance_name_2:
new_cap
})