1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
|
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You 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.
"""
Running workflows.
"""
import os
import sys
from datetime import datetime
from . import exceptions
from .context.workflow import WorkflowContext
from .workflows import builtin
from .workflows.core import engine, graph_compiler
from .workflows.executor.process import ProcessExecutor
from ..modeling import models
from ..modeling import utils as modeling_utils
from ..utils.imports import import_fullname
DEFAULT_TASK_MAX_ATTEMPTS = 30
DEFAULT_TASK_RETRY_INTERVAL = 30
class WorkflowRunner(object):
def __init__(self, model_storage, resource_storage, plugin_manager,
execution_id=None, retry_failed_tasks=False,
service_id=None, workflow_name=None, inputs=None, executor=None,
task_max_attempts=DEFAULT_TASK_MAX_ATTEMPTS,
task_retry_interval=DEFAULT_TASK_RETRY_INTERVAL):
"""
Manages a single workflow execution on a given service.
:param workflow_name: workflow name
:param service_id: service ID
:param inputs: key-value dict of inputs for the execution
:param model_storage: model storage API ("MAPI")
:param resource_storage: resource storage API ("RAPI")
:param plugin_manager: plugin manager
:param executor: executor for tasks; defaults to a
:class:`~aria.orchestrator.workflows.executor.process.ProcessExecutor` instance
:param task_max_attempts: maximum attempts of repeating each failing task
:param task_retry_interval: retry interval between retry attempts of a failing task
"""
if not (execution_id or (workflow_name and service_id)):
exceptions.InvalidWorkflowRunnerParams(
"Either provide execution id in order to resume a workflow or workflow name "
"and service id with inputs")
self._is_resume = execution_id is not None
self._retry_failed_tasks = retry_failed_tasks
self._model_storage = model_storage
self._resource_storage = resource_storage
# the IDs are stored rather than the models themselves, so this module could be used
# by several threads without raising errors on model objects shared between threadsF
if self._is_resume:
self._service_id = service_id
# self._service_id = self.execution.service.id
# self._workflow_name = model_storage.execution.get(self._execution_id).workflow_name
self._workflow_name = workflow_name
self._validate_workflow_exists_for_service()
self._execution_id = execution_id
else:
self._service_id = service_id
self._workflow_name = workflow_name
self._validate_workflow_exists_for_service()
self._execution_id = self._create_execution_model(inputs).id
self._create_execution_model(inputs, execution_id)
self._workflow_context = WorkflowContext(
name=self.__class__.__name__,
model_storage=self._model_storage,
resource_storage=resource_storage,
service_id=service_id,
execution_id=execution_id,
workflow_name=self._workflow_name,
task_max_attempts=task_max_attempts,
task_retry_interval=task_retry_interval)
# Set default executor and kwargs
executor = executor or ProcessExecutor(plugin_manager=plugin_manager)
# transforming the execution inputs to dict, to pass them to the workflow function
# execution_inputs_dict = dict(inp.unwrapped for inp in self.execution.inputs.itervalues())
# if not self._is_resume:
# workflow_fn = self._get_workflow_fn()
# self._tasks_graph = workflow_fn(ctx=self._workflow_context, **execution_inputs_dict)
# compiler = graph_compiler.GraphCompiler(self._workflow_context, executor.__class__)
# compiler.compile(self._tasks_graph)
self._engine = engine.Engine(executors={executor.__class__: executor})
@property
def execution_id(self):
return self._execution_id
@property
def execution(self):
return self._model_storage.execution.get(self._execution_id)
@property
def service(self):
return self._model_storage.service.get(self._service_id)
def execute(self):
self._engine.execute(ctx=self._workflow_context,
resuming=self._is_resume,
retry_failed=self._retry_failed_tasks)
def cancel(self):
self._engine.cancel_execution(ctx=self._workflow_context)
def _create_execution_model(self, inputs, execution_id):
execution = models.Execution(
created_at=datetime.utcnow(),
service=self.service,
workflow_name=self._workflow_name,
inputs={})
if self._workflow_name in builtin.BUILTIN_WORKFLOWS:
workflow_inputs = dict() # built-in workflows don't have any inputs
else:
workflow_inputs = self.service.workflows[self._workflow_name].inputs
# modeling_utils.validate_no_undeclared_inputs(declared_inputs=workflow_inputs,
# supplied_inputs=inputs or {})
modeling_utils.validate_required_inputs_are_supplied(declared_inputs=workflow_inputs,
supplied_inputs=inputs or {})
execution.inputs = modeling_utils.merge_parameter_values(
inputs, workflow_inputs, model_cls=models.Input)
execution.id = execution_id
# TODO: these two following calls should execute atomically
self._validate_no_active_executions(execution)
self._model_storage.execution.put(execution)
return execution
def _validate_workflow_exists_for_service(self):
if self._workflow_name not in self.service.workflows and \
self._workflow_name not in builtin.BUILTIN_WORKFLOWS:
raise exceptions.UndeclaredWorkflowError(
'No workflow policy {0} declared in service {1}'
.format(self._workflow_name, self.service.name))
def _validate_no_active_executions(self, execution):
active_executions = [e for e in self.service.executions if e.is_active()]
if active_executions:
raise exceptions.ActiveExecutionsError(
"Can't start execution; Service {0} has an active execution with ID {1}"
.format(self.service.name, active_executions[0].id))
def _get_workflow_fn(self):
if self._workflow_name in builtin.BUILTIN_WORKFLOWS:
return import_fullname('{0}.{1}'.format(builtin.BUILTIN_WORKFLOWS_PATH_PREFIX,
self._workflow_name))
workflow = self.service.workflows[self._workflow_name]
# TODO: Custom workflow support needs improvement, currently this code uses internal
# knowledge of the resource storage; Instead, workflows should probably be loaded
# in a similar manner to operation plugins. Also consider passing to import_fullname
# as paths instead of appending to sys path.
service_template_resources_path = os.path.join(
self._resource_storage.service_template.base_path,
str(self.service.service_template.id))
sys.path.append(service_template_resources_path)
try:
workflow_fn = import_fullname(workflow.function)
except ImportError:
raise exceptions.WorkflowImplementationNotFoundError(
'Could not find workflow {0} function at {1}'.format(
self._workflow_name, workflow.function))
return workflow_fn
|
