ansible/lib/ansible/executor/play_iterator.py
James Cammarata ca5b361ad8 Reworking iterator logic regarding failed states during always
Previous changes addressed a corner case, which unfortunately introduced
another bug. This patch adds a new flag to the host state (did_rescue) which
is set to true when the rescue portion of a block completes. This flag is
then checked in _check_failed_state() when the fail_state != FAILED_NONE.

This lead to the discovery of another bug - current strategies are not advancing
hosts to ITERATING_COMPLETE after doing a peek at the next task, leaving the
host state in the run_state of the final task. To address this, before gathering
the list of failed hosts in StrategyBase.run(), a final pass through the iterator
for all hosts is done to ensure each host is in its final state. This way, no
strategy derived from StrategyBase has to worry about it and it's handled.

Fixes #17983
2016-11-16 10:21:46 -06:00

596 lines
27 KiB
Python

# (c) 2012-2014, Michael DeHaan <michael.dehaan@gmail.com>
#
# This file is part of Ansible
#
# Ansible is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Ansible is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Ansible. If not, see <http://www.gnu.org/licenses/>.
# Make coding more python3-ish
from __future__ import (absolute_import, division, print_function)
__metaclass__ = type
import fnmatch
from ansible.compat.six import iteritems
from ansible import constants as C
from ansible.errors import AnsibleError
from ansible.playbook.block import Block
from ansible.playbook.task import Task
from ansible.playbook.role_include import IncludeRole
from ansible.utils.boolean import boolean
__all__ = ['PlayIterator']
try:
from __main__ import display
except ImportError:
from ansible.utils.display import Display
display = Display()
class HostState:
def __init__(self, blocks):
self._blocks = blocks[:]
self.cur_block = 0
self.cur_regular_task = 0
self.cur_rescue_task = 0
self.cur_always_task = 0
self.cur_role = None
self.cur_dep_chain = None
self.run_state = PlayIterator.ITERATING_SETUP
self.fail_state = PlayIterator.FAILED_NONE
self.pending_setup = False
self.tasks_child_state = None
self.rescue_child_state = None
self.always_child_state = None
self.did_rescue = False
self.did_start_at_task = False
def __repr__(self):
return "HostState(%r)" % self._blocks
def __str__(self):
def _run_state_to_string(n):
states = ["ITERATING_SETUP", "ITERATING_TASKS", "ITERATING_RESCUE", "ITERATING_ALWAYS", "ITERATING_COMPLETE"]
try:
return states[n]
except IndexError:
return "UNKNOWN STATE"
def _failed_state_to_string(n):
states = {1:"FAILED_SETUP", 2:"FAILED_TASKS", 4:"FAILED_RESCUE", 8:"FAILED_ALWAYS"}
if n == 0:
return "FAILED_NONE"
else:
ret = []
for i in (1, 2, 4, 8):
if n & i:
ret.append(states[i])
return "|".join(ret)
return "HOST STATE: block=%d, task=%d, rescue=%d, always=%d, role=%s, run_state=%s, fail_state=%s, pending_setup=%s, tasks child state? (%s), rescue child state? (%s), always child state? (%s), did rescue? %s, did start at task? %s" % (
self.cur_block,
self.cur_regular_task,
self.cur_rescue_task,
self.cur_always_task,
self.cur_role,
_run_state_to_string(self.run_state),
_failed_state_to_string(self.fail_state),
self.pending_setup,
self.tasks_child_state,
self.rescue_child_state,
self.always_child_state,
self.did_rescue,
self.did_start_at_task,
)
def __eq__(self, other):
if not isinstance(other, HostState):
return False
for attr in (
'_blocks', 'cur_block', 'cur_regular_task', 'cur_rescue_task', 'cur_always_task',
'cur_role', 'run_state', 'fail_state', 'pending_setup', 'cur_dep_chain',
'tasks_child_state', 'rescue_child_state', 'always_child_state'
):
if getattr(self, attr) != getattr(other, attr):
return False
return True
def get_current_block(self):
return self._blocks[self.cur_block]
def copy(self):
new_state = HostState(self._blocks)
new_state.cur_block = self.cur_block
new_state.cur_regular_task = self.cur_regular_task
new_state.cur_rescue_task = self.cur_rescue_task
new_state.cur_always_task = self.cur_always_task
new_state.cur_role = self.cur_role
new_state.run_state = self.run_state
new_state.fail_state = self.fail_state
new_state.pending_setup = self.pending_setup
new_state.did_rescue = self.did_rescue
new_state.did_start_at_task = self.did_start_at_task
if self.cur_dep_chain is not None:
new_state.cur_dep_chain = self.cur_dep_chain[:]
if self.tasks_child_state is not None:
new_state.tasks_child_state = self.tasks_child_state.copy()
if self.rescue_child_state is not None:
new_state.rescue_child_state = self.rescue_child_state.copy()
if self.always_child_state is not None:
new_state.always_child_state = self.always_child_state.copy()
return new_state
class PlayIterator:
# the primary running states for the play iteration
ITERATING_SETUP = 0
ITERATING_TASKS = 1
ITERATING_RESCUE = 2
ITERATING_ALWAYS = 3
ITERATING_COMPLETE = 4
# the failure states for the play iteration, which are powers
# of 2 as they may be or'ed together in certain circumstances
FAILED_NONE = 0
FAILED_SETUP = 1
FAILED_TASKS = 2
FAILED_RESCUE = 4
FAILED_ALWAYS = 8
def __init__(self, inventory, play, play_context, variable_manager, all_vars, start_at_done=False):
self._play = play
self._blocks = []
self._task_uuid_cache = dict()
# Default options to gather
gather_subset = C.DEFAULT_GATHER_SUBSET
gather_timeout = C.DEFAULT_GATHER_TIMEOUT
# Retrieve subset to gather
if self._play.gather_subset is not None:
gather_subset = self._play.gather_subset
# Retrieve timeout for gather
if self._play.gather_timeout is not None:
gather_timeout = self._play.gather_timeout
setup_block = Block(play=self._play)
setup_task = Task(block=setup_block)
setup_task.action = 'setup'
setup_task.name = 'Gathering Facts'
setup_task.tags = ['always']
setup_task.args = {
'gather_subset': gather_subset,
}
if gather_timeout:
setup_task.args['gather_timeout'] = gather_timeout
setup_task.set_loader(self._play._loader)
setup_block.block = [setup_task]
setup_block = setup_block.filter_tagged_tasks(play_context, all_vars)
self._blocks.append(setup_block)
self.cache_block_tasks(setup_block)
for block in self._play.compile():
new_block = block.filter_tagged_tasks(play_context, all_vars)
if new_block.has_tasks():
self.cache_block_tasks(new_block)
self._blocks.append(new_block)
for handler_block in self._play.handlers:
self.cache_block_tasks(handler_block)
self._host_states = {}
start_at_matched = False
for host in inventory.get_hosts(self._play.hosts):
self._host_states[host.name] = HostState(blocks=self._blocks)
# if the host's name is in the variable manager's fact cache, then set
# its _gathered_facts flag to true for smart gathering tests later
if host.name in variable_manager._fact_cache and variable_manager._fact_cache.get('module_setup', False):
host._gathered_facts = True
# if we're looking to start at a specific task, iterate through
# the tasks for this host until we find the specified task
if play_context.start_at_task is not None and not start_at_done:
while True:
(s, task) = self.get_next_task_for_host(host, peek=True)
if s.run_state == self.ITERATING_COMPLETE:
break
if task.name == play_context.start_at_task or fnmatch.fnmatch(task.name, play_context.start_at_task) or \
task.get_name() == play_context.start_at_task or fnmatch.fnmatch(task.get_name(), play_context.start_at_task):
start_at_matched = True
break
else:
self.get_next_task_for_host(host)
# finally, reset the host's state to ITERATING_SETUP
if start_at_matched:
self._host_states[host.name].did_start_at_task = True
self._host_states[host.name].run_state = self.ITERATING_SETUP
if start_at_matched:
# we have our match, so clear the start_at_task field on the
# play context to flag that we've started at a task (and future
# plays won't try to advance)
play_context.start_at_task = None
def get_host_state(self, host):
# Since we're using the PlayIterator to carry forward failed hosts,
# in the event that a previous host was not in the current inventory
# we create a stub state for it now
if host.name not in self._host_states:
self._host_states[host.name] = HostState(blocks=[])
return self._host_states[host.name].copy()
def cache_block_tasks(self, block):
def _cache_portion(p):
for t in p:
if isinstance(t, Block):
self.cache_block_tasks(t)
elif t._uuid not in self._task_uuid_cache:
self._task_uuid_cache[t._uuid] = t
for portion in (block.block, block.rescue, block.always):
if portion is not None:
_cache_portion(portion)
def get_next_task_for_host(self, host, peek=False):
display.debug("getting the next task for host %s" % host.name)
s = self.get_host_state(host)
task = None
if s.run_state == self.ITERATING_COMPLETE:
display.debug("host %s is done iterating, returning" % host.name)
return (s, None)
old_s = s
(s, task) = self._get_next_task_from_state(s, host=host, peek=peek)
def _roles_are_different(ra, rb):
if ra != rb:
return True
else:
return old_s.cur_dep_chain != task.get_dep_chain()
def _role_is_child(r):
parent = task._parent
while parent:
if hasattr(parent, '_role') and parent._role == r and isinstance(parent, IncludeRole):
return True
parent = parent._parent
return False
if task and task._role:
# if we had a current role, mark that role as completed
if s.cur_role and _roles_are_different(task._role, s.cur_role) and host.name in s.cur_role._had_task_run and \
not _role_is_child(s.cur_role) and not peek:
s.cur_role._completed[host.name] = True
s.cur_role = task._role
s.cur_dep_chain = task.get_dep_chain()
if not peek:
self._host_states[host.name] = s
display.debug("done getting next task for host %s" % host.name)
display.debug(" ^ task is: %s" % task)
display.debug(" ^ state is: %s" % s)
return (s, task)
def _get_next_task_from_state(self, state, host, peek):
task = None
# try and find the next task, given the current state.
while True:
# try to get the current block from the list of blocks, and
# if we run past the end of the list we know we're done with
# this block
try:
block = state._blocks[state.cur_block]
except IndexError:
state.run_state = self.ITERATING_COMPLETE
return (state, None)
if state.run_state == self.ITERATING_SETUP:
# First, we check to see if we were pending setup. If not, this is
# the first trip through ITERATING_SETUP, so we set the pending_setup
# flag and try to determine if we do in fact want to gather facts for
# the specified host.
if not state.pending_setup:
state.pending_setup = True
# Gather facts if the default is 'smart' and we have not yet
# done it for this host; or if 'explicit' and the play sets
# gather_facts to True; or if 'implicit' and the play does
# NOT explicitly set gather_facts to False.
gathering = C.DEFAULT_GATHERING
implied = self._play.gather_facts is None or boolean(self._play.gather_facts)
if (gathering == 'implicit' and implied) or \
(gathering == 'explicit' and boolean(self._play.gather_facts)) or \
(gathering == 'smart' and implied and not host._gathered_facts):
# The setup block is always self._blocks[0], as we inject it
# during the play compilation in __init__ above.
setup_block = self._blocks[0]
if setup_block.has_tasks() and len(setup_block.block) > 0:
task = setup_block.block[0]
if not peek:
# mark the host as having gathered facts, because we're
# returning the setup task to be executed
host.set_gathered_facts(True)
else:
# This is the second trip through ITERATING_SETUP, so we clear
# the flag and move onto the next block in the list while setting
# the run state to ITERATING_TASKS
state.pending_setup = False
state.run_state = self.ITERATING_TASKS
if not state.did_start_at_task:
state.cur_block += 1
state.cur_regular_task = 0
state.cur_rescue_task = 0
state.cur_always_task = 0
state.child_state = None
elif state.run_state == self.ITERATING_TASKS:
# clear the pending setup flag, since we're past that and it didn't fail
if state.pending_setup:
state.pending_setup = False
# First, we check for a child task state that is not failed, and if we
# have one recurse into it for the next task. If we're done with the child
# state, we clear it and drop back to geting the next task from the list.
if state.tasks_child_state:
(state.tasks_child_state, task) = self._get_next_task_from_state(state.tasks_child_state, host=host, peek=peek)
if self._check_failed_state(state.tasks_child_state):
# failed child state, so clear it and move into the rescue portion
state.tasks_child_state = None
self._set_failed_state(state)
else:
# get the next task recursively
if task is None or state.tasks_child_state.run_state == self.ITERATING_COMPLETE:
# we're done with the child state, so clear it and continue
# back to the top of the loop to get the next task
state.tasks_child_state = None
continue
else:
# First here, we check to see if we've failed anywhere down the chain
# of states we have, and if so we move onto the rescue portion. Otherwise,
# we check to see if we've moved past the end of the list of tasks. If so,
# we move into the always portion of the block, otherwise we get the next
# task from the list.
if self._check_failed_state(state):
state.run_state = self.ITERATING_RESCUE
elif state.cur_regular_task >= len(block.block):
state.run_state = self.ITERATING_ALWAYS
else:
task = block.block[state.cur_regular_task]
# if the current task is actually a child block, create a child
# state for us to recurse into on the next pass
if isinstance(task, Block) or state.tasks_child_state is not None:
state.tasks_child_state = HostState(blocks=[task])
state.tasks_child_state.run_state = self.ITERATING_TASKS
state.tasks_child_state.cur_role = state.cur_role
# since we've created the child state, clear the task
# so we can pick up the child state on the next pass
task = None
state.cur_regular_task += 1
elif state.run_state == self.ITERATING_RESCUE:
# The process here is identical to ITERATING_TASKS, except instead
# we move into the always portion of the block.
if state.rescue_child_state:
(state.rescue_child_state, task) = self._get_next_task_from_state(state.rescue_child_state, host=host, peek=peek)
if self._check_failed_state(state.rescue_child_state):
state.rescue_child_state = None
self._set_failed_state(state)
else:
if task is None or state.rescue_child_state.run_state == self.ITERATING_COMPLETE:
state.rescue_child_state = None
continue
else:
if state.fail_state & self.FAILED_RESCUE == self.FAILED_RESCUE:
state.run_state = self.ITERATING_ALWAYS
elif state.cur_rescue_task >= len(block.rescue):
if len(block.rescue) > 0:
state.fail_state = self.FAILED_NONE
state.run_state = self.ITERATING_ALWAYS
state.did_rescue = True
else:
task = block.rescue[state.cur_rescue_task]
if isinstance(task, Block) or state.rescue_child_state is not None:
state.rescue_child_state = HostState(blocks=[task])
state.rescue_child_state.run_state = self.ITERATING_TASKS
state.rescue_child_state.cur_role = state.cur_role
task = None
state.cur_rescue_task += 1
elif state.run_state == self.ITERATING_ALWAYS:
# And again, the process here is identical to ITERATING_TASKS, except
# instead we either move onto the next block in the list, or we set the
# run state to ITERATING_COMPLETE in the event of any errors, or when we
# have hit the end of the list of blocks.
if state.always_child_state:
(state.always_child_state, task) = self._get_next_task_from_state(state.always_child_state, host=host, peek=peek)
if self._check_failed_state(state.always_child_state):
state.always_child_state = None
self._set_failed_state(state)
else:
if task is None or state.always_child_state.run_state == self.ITERATING_COMPLETE:
state.always_child_state = None
continue
else:
if state.cur_always_task >= len(block.always):
if state.fail_state != self.FAILED_NONE:
state.run_state = self.ITERATING_COMPLETE
else:
state.cur_block += 1
state.cur_regular_task = 0
state.cur_rescue_task = 0
state.cur_always_task = 0
state.run_state = self.ITERATING_TASKS
state.tasks_child_state = None
state.rescue_child_state = None
state.always_child_state = None
state.did_rescue = False
else:
task = block.always[state.cur_always_task]
if isinstance(task, Block) or state.always_child_state is not None:
state.always_child_state = HostState(blocks=[task])
state.always_child_state.run_state = self.ITERATING_TASKS
state.always_child_state.cur_role = state.cur_role
task = None
state.cur_always_task += 1
elif state.run_state == self.ITERATING_COMPLETE:
return (state, None)
# if something above set the task, break out of the loop now
if task:
break
return (state, task)
def _set_failed_state(self, state):
if state.run_state == self.ITERATING_SETUP:
state.fail_state |= self.FAILED_SETUP
state.run_state = self.ITERATING_COMPLETE
elif state.run_state == self.ITERATING_TASKS:
if state.tasks_child_state is not None:
state.tasks_child_state = self._set_failed_state(state.tasks_child_state)
else:
state.fail_state |= self.FAILED_TASKS
if state._blocks[state.cur_block].rescue:
state.run_state = self.ITERATING_RESCUE
elif state._blocks[state.cur_block].always:
state.run_state = self.ITERATING_ALWAYS
else:
state.run_state = self.ITERATING_COMPLETE
elif state.run_state == self.ITERATING_RESCUE:
if state.rescue_child_state is not None:
state.rescue_child_state = self._set_failed_state(state.rescue_child_state)
else:
state.fail_state |= self.FAILED_RESCUE
if state._blocks[state.cur_block].always:
state.run_state = self.ITERATING_ALWAYS
else:
state.run_state = self.ITERATING_COMPLETE
elif state.run_state == self.ITERATING_ALWAYS:
if state.always_child_state is not None:
state.always_child_state = self._set_failed_state(state.always_child_state)
else:
state.fail_state |= self.FAILED_ALWAYS
state.run_state = self.ITERATING_COMPLETE
return state
def mark_host_failed(self, host):
s = self.get_host_state(host)
display.debug("marking host %s failed, current state: %s" % (host, s))
s = self._set_failed_state(s)
display.debug("^ failed state is now: %s" % s)
self._host_states[host.name] = s
self._play._removed_hosts.append(host.name)
def get_failed_hosts(self):
return dict((host, True) for (host, state) in iteritems(self._host_states) if self._check_failed_state(state))
def _check_failed_state(self, state):
if state is None:
return False
elif state.run_state == self.ITERATING_RESCUE and self._check_failed_state(state.rescue_child_state):
return True
elif state.run_state == self.ITERATING_ALWAYS and self._check_failed_state(state.always_child_state):
return True
elif state.fail_state != self.FAILED_NONE:
if state.run_state == self.ITERATING_RESCUE and state.fail_state&self.FAILED_RESCUE == 0:
return False
elif state.run_state == self.ITERATING_ALWAYS and state.fail_state&self.FAILED_ALWAYS == 0:
return False
else:
return not state.did_rescue
elif state.run_state == self.ITERATING_TASKS and self._check_failed_state(state.tasks_child_state):
cur_block = self._blocks[state.cur_block]
if len(cur_block.rescue) > 0 and state.fail_state & self.FAILED_RESCUE == 0:
return False
else:
return True
return False
def is_failed(self, host):
s = self.get_host_state(host)
return self._check_failed_state(s)
def get_original_task(self, host, task):
'''
Finds the task in the task list which matches the UUID of the given task.
The executor engine serializes/deserializes objects as they are passed through
the different processes, and not all data structures are preserved. This method
allows us to find the original task passed into the executor engine.
'''
if isinstance(task, Task):
the_uuid = task._uuid
else:
the_uuid = task
return self._task_uuid_cache.get(the_uuid, None)
def _insert_tasks_into_state(self, state, task_list):
# if we've failed at all, or if the task list is empty, just return the current state
if state.fail_state != self.FAILED_NONE and state.run_state not in (self.ITERATING_RESCUE, self.ITERATING_ALWAYS) or not task_list:
return state
if state.run_state == self.ITERATING_TASKS:
if state.tasks_child_state:
state.tasks_child_state = self._insert_tasks_into_state(state.tasks_child_state, task_list)
else:
target_block = state._blocks[state.cur_block].copy(exclude_parent=True)
before = target_block.block[:state.cur_regular_task]
after = target_block.block[state.cur_regular_task:]
target_block.block = before + task_list + after
state._blocks[state.cur_block] = target_block
elif state.run_state == self.ITERATING_RESCUE:
if state.rescue_child_state:
state.rescue_child_state = self._insert_tasks_into_state(state.rescue_child_state, task_list)
else:
target_block = state._blocks[state.cur_block].copy(exclude_parent=True)
before = target_block.rescue[:state.cur_rescue_task]
after = target_block.rescue[state.cur_rescue_task:]
target_block.rescue = before + task_list + after
state._blocks[state.cur_block] = target_block
elif state.run_state == self.ITERATING_ALWAYS:
if state.always_child_state:
state.always_child_state = self._insert_tasks_into_state(state.always_child_state, task_list)
else:
target_block = state._blocks[state.cur_block].copy(exclude_parent=True)
before = target_block.always[:state.cur_always_task]
after = target_block.always[state.cur_always_task:]
target_block.always = before + task_list + after
state._blocks[state.cur_block] = target_block
return state
def add_tasks(self, host, task_list):
for b in task_list:
self.cache_block_tasks(b)
self._host_states[host.name] = self._insert_tasks_into_state(self.get_host_state(host), task_list)