2 # NEPI, a framework to manage network experiments
3 # Copyright (C) 2013 INRIA
5 # This program is free software: you can redistribute it and/or modify
6 # it under the terms of the GNU General Public License as published by
7 # the Free Software Foundation, either version 3 of the License, or
8 # (at your option) any later version.
10 # This program is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
15 # You should have received a copy of the GNU General Public License
16 # along with this program. If not, see <http://www.gnu.org/licenses/>.
18 # Author: Alina Quereilhac <alina.quereilhac@inria.fr>
20 from nepi.util import guid
21 from nepi.util.parallel import ParallelRun
22 from nepi.util.timefuncs import tnow, tdiffsec, stabsformat, tsformat
23 from nepi.execution.resource import ResourceFactory, ResourceAction, \
24 ResourceState, ResourceState2str
25 from nepi.execution.scheduler import HeapScheduler, Task, TaskStatus
26 from nepi.execution.trace import TraceAttr
28 # TODO: use multiprocessing instead of threading
29 # TODO: Allow to reconnect to a running experiment instance! (reconnect mode vs deploy mode)
40 class FailureLevel(object):
41 """ Describes the system failure state
47 class FailureManager(object):
48 """ The FailureManager is responsible for handling errors,
49 and deciding whether an experiment should be aborted
52 def __init__(self, ec):
53 self._ec = weakref.ref(ec)
54 self._failure_level = FailureLevel.OK
58 """ Returns the Experiment Controller """
63 if self._failure_level == FailureLevel.OK:
64 for guid in self.ec.resources:
65 state = self.ec.state(guid)
66 critical = self.ec.get(guid, "critical")
68 if state == ResourceState.FAILED and critical:
69 self._failure_level = FailureLevel.RM_FAILURE
70 self.ec.logger.debug("RM critical failure occurred on guid %d." \
71 " Setting EC FAILURE LEVEL to RM_FAILURE" % guid)
74 return self._failure_level != FailureLevel.OK
76 def set_ec_failure(self):
77 self._failure_level = FailureLevel.EC_FAILURE
80 class ECState(object):
81 """ State of the Experiment Controller
88 class ExperimentController(object):
90 .. class:: Class Args :
92 :param exp_id: Human readable identifier for the experiment scenario.
97 An experiment, or scenario, is defined by a concrete set of resources,
98 behavior, configuration and interconnection of those resources.
99 The Experiment Description (ED) is a detailed representation of a
100 single experiment. It contains all the necessary information to
101 allow repeating the experiment. NEPI allows to describe
102 experiments by registering components (resources), configuring them
103 and interconnecting them.
105 A same experiment (scenario) can be executed many times, generating
106 different results. We call an experiment execution (instance) a 'run'.
108 The ExperimentController (EC), is the entity responsible of
109 managing an experiment run. The same scenario can be
110 recreated (and re-run) by instantiating an EC and recreating
111 the same experiment description.
113 In NEPI, an experiment is represented as a graph of interconnected
114 resources. A resource is a generic concept in the sense that any
115 component taking part of an experiment, whether physical of
116 virtual, is considered a resource. A resources could be a host,
117 a virtual machine, an application, a simulator, a IP address.
119 A ResourceManager (RM), is the entity responsible for managing a
120 single resource. ResourceManagers are specific to a resource
121 type (i.e. An RM to control a Linux application will not be
122 the same as the RM used to control a ns-3 simulation).
123 To support a new type of resource in NEPI, a new RM must be
124 implemented. NEPI already provides a variety of
125 RMs to control basic resources, and new can be extended from
128 Through the EC interface the user can create ResourceManagers (RMs),
129 configure them and interconnect them, to describe an experiment.
130 Describing an experiment through the EC does not run the experiment.
131 Only when the 'deploy()' method is invoked on the EC, the EC will take
132 actions to transform the 'described' experiment into a 'running' experiment.
134 While the experiment is running, it is possible to continue to
135 create/configure/connect RMs, and to deploy them to involve new
136 resources in the experiment (this is known as 'interactive' deployment).
138 An experiments in NEPI is identified by a string id,
139 which is either given by the user, or automatically generated by NEPI.
140 The purpose of this identifier is to separate files and results that
141 belong to different experiment scenarios.
142 However, since a same 'experiment' can be run many times, the experiment
143 id is not enough to identify an experiment instance (run).
144 For this reason, the ExperimentController has two identifier, the
145 exp_id, which can be re-used in different ExperimentController,
146 and the run_id, which is unique to one ExperimentController instance, and
147 is automatically generated by NEPI.
151 def __init__(self, exp_id = None):
152 super(ExperimentController, self).__init__()
154 self._logger = logging.getLogger("ExperimentController")
156 # Run identifier. It identifies a concrete execution instance (run)
158 # Since a same experiment (same configuration) can be executed many
159 # times, this run_id permits to separate result files generated on
160 # different experiment executions
161 self._run_id = tsformat()
163 # Experiment identifier. Usually assigned by the user
164 # Identifies the experiment scenario (i.e. configuration,
165 # resources used, etc)
166 self._exp_id = exp_id or "exp-%s" % os.urandom(8).encode('hex')
168 # generator of globally unique ids
169 self._guid_generator = guid.GuidGenerator()
172 self._resources = dict()
174 # Scheduler. It a queue that holds tasks scheduled for
175 # execution, and yields the next task to be executed
176 # ordered by execution and arrival time
177 self._scheduler = HeapScheduler()
182 # RM groups (for deployment)
183 self._groups = dict()
185 # generator of globally unique id for groups
186 self._group_id_generator = guid.GuidGenerator()
188 # Flag to stop processing thread
191 # Entity in charge of managing system failures
192 self._fm = FailureManager(self)
195 self._state = ECState.RUNNING
197 # The runner is a pool of threads used to parallelize
199 nthreads = int(os.environ.get("NEPI_NTHREADS", "50"))
200 self._runner = ParallelRun(maxthreads = nthreads)
202 # Event processing thread
203 self._cond = threading.Condition()
204 self._thread = threading.Thread(target = self._process)
205 self._thread.setDaemon(True)
210 """ Return the logger of the Experiment Controller
217 """ Return the state of the Experiment Controller
224 """ Return the experiment id assigned by the user
231 """ Return the experiment instance (run) identifier
238 return self._fm.abort
240 def wait_finished(self, guids):
241 """ Blocking method that wait until all RMs in the 'guid' list
242 reach a state >= STOPPED (i.e. FINISHED, STOPPED, FAILED or
243 RELEASED ) or until a System Failure occurs (e.g. Task Failure)
245 :param guids: List of guids
253 return self.wait(guids, state = ResourceState.STOPPED,
256 def wait_started(self, guids):
257 """ Blocking method that wait until all RMs in the 'guid' list
258 reach a state >= STARTED or until a System Failure occurs
261 :param guids: List of guids
268 return self.wait(guids, state = ResourceState.STARTED,
271 def wait_released(self, guids):
272 """ Blocking method that wait until all RMs in the 'guid' list
273 reach a state = RELEASED or until the EC fails
275 :param guids: List of guids
280 return self._state == ECState.FAILED
282 return self.wait(guids, state = ResourceState.RELEASED,
285 def wait_deployed(self, guids):
286 """ Blocking method that wait until all RMs in the 'guid' list
287 reach a state >= READY or until a System Failure occurs
290 :param guids: List of guids
297 return self.wait(guids, state = ResourceState.READY,
300 def wait(self, guids, state, quit):
301 """ Blocking method that wait until all RMs in the 'guid' list
302 reach a state >= 'state' or until quit yileds True
304 :param guids: List of guids
307 if isinstance(guids, int):
310 # Make a copy to avoid modifying the original guids list
314 # If there are no more guids to wait for
315 # or the quit function returns True, exit the loop
316 if len(guids) == 0 or quit():
319 # If a guid reached one of the target states, remove it from list
321 rstate = self.state(guid)
323 hrrstate = ResourceState2str.get(rstate)
324 hrstate = ResourceState2str.get(state)
328 rm = self.get_resource(guid)
329 self.logger.debug(" %s guid %d DONE - state is %s, required is >= %s " % (
330 rm.rtype(), guid, hrrstate, hrstate))
333 self.logger.debug(" WAITING FOR guid %d - state is %s, required is >= %s " % (
334 guid, hrrstate, hrstate))
337 def get_task(self, tid):
338 """ Get a specific task
340 :param tid: Id of the task
344 return self._tasks.get(tid)
346 def get_resource(self, guid):
347 """ Get a specific Resource Manager
349 :param guid: Id of the task
351 :rtype: ResourceManager
353 return self._resources.get(guid)
357 """ Returns the list of all the Resource Manager Id
362 return self._resources.keys()
364 def register_resource(self, rtype, guid = None):
365 """ Register a Resource Manager. It creates a new 'guid', if it is not specified,
366 for the RM of type 'rtype' and add it to the list of Resources.
368 :param rtype: Type of the RM
370 :return: Id of the RM
373 # Get next available guid
374 guid = self._guid_generator.next(guid)
377 rm = ResourceFactory.create(rtype, self, guid)
380 self._resources[guid] = rm
384 def get_attributes(self, guid):
385 """ Return all the attibutes of a specific RM
387 :param guid: Guid of the RM
389 :return: List of attributes
392 rm = self.get_resource(guid)
393 return rm.get_attributes()
395 def register_connection(self, guid1, guid2):
396 """ Registers a guid1 with a guid2.
397 The declaration order is not important
399 :param guid1: First guid to connect
400 :type guid1: ResourceManager
402 :param guid2: Second guid to connect
403 :type guid: ResourceManager
405 rm1 = self.get_resource(guid1)
406 rm2 = self.get_resource(guid2)
408 rm1.register_connection(guid2)
409 rm2.register_connection(guid1)
411 def register_condition(self, guids1, action, guids2, state,
413 """ Registers an action START or STOP for all RM on guids1 to occur
414 time 'time' after all elements in guids2 reached state 'state'.
416 :param guids1: List of guids of RMs subjected to action
419 :param action: Action to register (either START or STOP)
420 :type action: ResourceAction
422 :param guids2: List of guids of RMs to we waited for
425 :param state: State to wait for on RMs (STARTED, STOPPED, etc)
426 :type state: ResourceState
428 :param time: Time to wait after guids2 has reached status
432 if isinstance(guids1, int):
434 if isinstance(guids2, int):
438 rm = self.get_resource(guid1)
439 rm.register_condition(action, guids2, state, time)
441 def enable_trace(self, guid, name):
444 :param name: Name of the trace
447 rm = self.get_resource(guid)
448 rm.enable_trace(name)
450 def trace_enabled(self, guid, name):
451 """ Returns True if trace is enabled
453 :param name: Name of the trace
456 rm = self.get_resource(guid)
457 return rm.trace_enabled(name)
459 def trace(self, guid, name, attr = TraceAttr.ALL, block = 512, offset = 0):
460 """ Get information on collected trace
462 :param name: Name of the trace
465 :param attr: Can be one of:
466 - TraceAttr.ALL (complete trace content),
467 - TraceAttr.STREAM (block in bytes to read starting at offset),
468 - TraceAttr.PATH (full path to the trace file),
469 - TraceAttr.SIZE (size of trace file).
472 :param block: Number of bytes to retrieve from trace, when attr is TraceAttr.STREAM
475 :param offset: Number of 'blocks' to skip, when attr is TraceAttr.STREAM
480 rm = self.get_resource(guid)
481 return rm.trace(name, attr, block, offset)
483 def discover(self, guid):
484 """ Discover a specific RM defined by its 'guid'
486 :param guid: Guid of the RM
490 rm = self.get_resource(guid)
493 def provision(self, guid):
494 """ Provision a specific RM defined by its 'guid'
496 :param guid: Guid of the RM
500 rm = self.get_resource(guid)
501 return rm.provision()
503 def get(self, guid, name):
504 """ Get a specific attribute 'name' from the RM 'guid'
506 :param guid: Guid of the RM
509 :param name: attribute's name
513 rm = self.get_resource(guid)
516 def set(self, guid, name, value):
517 """ Set a specific attribute 'name' from the RM 'guid'
518 with the value 'value'
520 :param guid: Guid of the RM
523 :param name: attribute's name
526 :param value: attribute's value
529 rm = self.get_resource(guid)
530 return rm.set(name, value)
532 def state(self, guid, hr = False):
533 """ Returns the state of a resource
535 :param guid: Resource guid
538 :param hr: Human readable. Forces return of a
539 status string instead of a number
543 rm = self.get_resource(guid)
547 return ResourceState2str.get(state)
551 def stop(self, guid):
552 """ Stop a specific RM defined by its 'guid'
554 :param guid: Guid of the RM
558 rm = self.get_resource(guid)
561 def start(self, guid):
562 """ Start a specific RM defined by its 'guid'
564 :param guid: Guid of the RM
568 rm = self.get_resource(guid)
571 def set_with_conditions(self, name, value, guids1, guids2, state,
573 """ Set value 'value' on attribute with name 'name' on all RMs of
574 guids1 when 'time' has elapsed since all elements in guids2
575 have reached state 'state'.
577 :param name: Name of attribute to set in RM
580 :param value: Value of attribute to set in RM
583 :param guids1: List of guids of RMs subjected to action
586 :param action: Action to register (either START or STOP)
587 :type action: ResourceAction
589 :param guids2: List of guids of RMs to we waited for
592 :param state: State to wait for on RMs (STARTED, STOPPED, etc)
593 :type state: ResourceState
595 :param time: Time to wait after guids2 has reached status
599 if isinstance(guids1, int):
601 if isinstance(guids2, int):
605 rm = self.get_resource(guid)
606 rm.set_with_conditions(name, value, guids2, state, time)
608 def deploy(self, guids = None, wait_all_ready = True, group = None):
609 """ Deploy all resource manager in guids list
611 :param guids: List of guids of RMs to deploy
614 :param wait_all_ready: Wait until all RMs are ready in
615 order to start the RMs
618 :param group: Id of deployment group in which to deploy RMs
622 self.logger.debug(" ------- DEPLOY START ------ ")
625 # If no guids list was passed, all 'NEW' RMs will be deployed
627 for guid in self.resources:
628 if self.state(guid) == ResourceState.NEW:
631 if isinstance(guids, int):
634 # Create deployment group
635 # New guids can be added to a same deployment group later on
639 group = self._group_id_generator.next()
641 if group not in self._groups:
642 self._groups[group] = []
644 self._groups[group].extend(guids)
646 def wait_all_and_start(group):
647 # Function that checks if all resources are READY
648 # before scheduling a start_with_conditions for each RM
651 # Get all guids in group
652 guids = self._groups[group]
655 if self.state(guid) < ResourceState.READY:
660 callback = functools.partial(wait_all_and_start, group)
661 self.schedule("1s", callback)
663 # If all resources are ready, we schedule the start
665 rm = self.get_resource(guid)
666 self.schedule("0s", rm.start_with_conditions)
668 if wait_all_ready and new_group:
669 # Schedule a function to check that all resources are
670 # READY, and only then schedule the start.
671 # This aims at reducing the number of tasks looping in the
673 # Instead of having many start tasks, we will have only one for
675 callback = functools.partial(wait_all_and_start, group)
676 self.schedule("0s", callback)
679 rm = self.get_resource(guid)
680 rm.deployment_group = group
681 self.schedule("0s", rm.deploy_with_conditions)
683 if not wait_all_ready:
684 self.schedule("0s", rm.start_with_conditions)
686 if rm.conditions.get(ResourceAction.STOP):
687 # Only if the RM has STOP conditions we
688 # schedule a stop. Otherwise the RM will stop immediately
689 self.schedule("0s", rm.stop_with_conditions)
691 def release(self, guids = None):
692 """ Release al RMs on the guids list or
693 all the resources if no list is specified
695 :param guids: List of RM guids
700 guids = self.resources
702 # Remove all pending tasks from the scheduler queue
703 for tid in list(self._scheduler.pending):
704 self._scheduler.remove(tid)
709 rm = self.get_resource(guid)
710 self.schedule("0s", rm.release)
712 self.wait_released(guids)
715 """ Shutdown the Experiment Controller.
716 Releases all the resources and stops task processing thread
719 # If there was a major failure we can't exit gracefully
720 if self._state == ECState.FAILED:
721 raise RuntimeError("EC failure. Can not exit gracefully")
725 # Mark the EC state as TERMINATED
726 self._state = ECState.TERMINATED
728 # Stop processing thread
731 # Notify condition to wake up the processing thread
734 if self._thread.is_alive():
737 def schedule(self, date, callback, track = False):
738 """ Schedule a callback to be executed at time date.
740 :param date: string containing execution time for the task.
741 It can be expressed as an absolute time, using
742 timestamp format, or as a relative time matching
743 ^\d+.\d+(h|m|s|ms|us)$
745 :param callback: code to be executed for the task. Must be a
746 Python function, and receives args and kwargs
749 :param track: if set to True, the task will be retrivable with
750 the get_task() method
752 :return : The Id of the task
754 timestamp = stabsformat(date)
755 task = Task(timestamp, callback)
756 task = self._scheduler.schedule(task)
759 self._tasks[task.id] = task
761 # Notify condition to wake up the processing thread
767 """ Process scheduled tasks.
771 The _process method is executed in an independent thread held by the
772 ExperimentController for as long as the experiment is running.
774 Tasks are scheduled by invoking the schedule method with a target callback.
775 The schedule method is given a execution time which controls the
776 order in which tasks are processed.
778 Tasks are processed in parallel using multithreading.
779 The environmental variable NEPI_NTHREADS can be used to control
780 the number of threads used to process tasks. The default value is 50.
784 To execute tasks in parallel, an ParallelRunner (PR) object, holding
785 a pool of threads (workers), is used.
786 For each available thread in the PR, the next task popped from
787 the scheduler queue is 'put' in the PR.
788 Upon receiving a task to execute, each PR worker (thread) invokes the
789 _execute method of the EC, passing the task as argument.
790 This method, calls task.callback inside a try/except block. If an
791 exception is raised by the tasks.callback, it will be trapped by the
792 try block, logged to standard error (usually the console), and the EC
793 state will be set to ECState.FAILED.
794 The invocation of _notify immediately after, forces the processing
795 loop in the _process method, to wake up if it was blocked waiting for new
796 tasks to arrived, and to check the EC state.
797 As the EC is in FAILED state, the processing loop exits and the
798 'finally' block is invoked. In the 'finally' block, the 'sync' method
799 of the PR is invoked, which forces the PR to raise any unchecked errors
800 that might have been raised by the workers.
806 while not self._stop:
810 task = self._scheduler.next()
813 # No task to execute. Wait for a new task to be scheduled.
816 # The task timestamp is in the future. Wait for timeout
817 # or until another task is scheduled.
819 if now < task.timestamp:
820 # Calculate timeout in seconds
821 timeout = tdiffsec(task.timestamp, now)
823 # Re-schedule task with the same timestamp
824 self._scheduler.schedule(task)
828 # Wait timeout or until a new task awakes the condition
829 self._cond.wait(timeout)
834 # Process tasks in parallel
835 self._runner.put(self._execute, task)
838 err = traceback.format_exc()
839 self.logger.error("Error while processing tasks in the EC: %s" % err)
841 # Set the EC to FAILED state
842 self._state = ECState.FAILED
844 # Set the FailureManager failure level to EC failure
845 self._fm.set_ec_failure()
847 self.logger.debug("Exiting the task processing loop ... ")
850 self._runner.destroy()
852 def _execute(self, task):
853 """ Executes a single task.
855 :param task: Object containing the callback to execute
860 If the invokation of the task callback raises an
861 exception, the processing thread of the ExperimentController
862 will be stopped and the experiment will be aborted.
866 task.status = TaskStatus.DONE
869 task.result = task.callback()
872 err = traceback.format_exc()
874 task.status = TaskStatus.ERROR
876 self.logger.error("Error occurred while executing task: %s" % err)
879 """ Awakes the processing thread in case it is blocked waiting
880 for a new task to be scheduled.