tripleo-common/tripleo_common/_stack_update.py
Brad P. Crochet b14c0f727e Remove update abort
The update abort is not reliable, and could mess up a TripleO stack
beyond repair. Since this has a potential for data loss, I would
suggest this be removed without a deprecation period.

Depends-On: Ieec4f01e38768eafb3df1f06340bdd3e220d30bd

Change-Id: I752e061979d667c1fb2b115c1a7339002e1824d5
2017-05-03 11:21:14 +00:00

204 lines
7.7 KiB
Python

# Copyright 2015 Red Hat, Inc.
# All Rights Reserved.
#
# 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.
import fnmatch
import logging
import re
import time
import six
import heatclient.exc
LOG = logging.getLogger(__name__)
class StackUpdateManager(object):
def __init__(self, heatclient, novaclient, stack, hook_type,
nested_depth=5, hook_resource=None):
self.heatclient = heatclient
self.novaclient = novaclient
self.stack = stack
self.hook_type = hook_type
self.nested_depth = nested_depth
self.hook_resource = hook_resource
self.server_names = {}
self.servers = []
def clear_breakpoints(self, refs):
resources = self._resources_by_state()
succeeds = []
fails = []
for ref in refs:
server_name = None
try:
res = resources['on_breakpoint'][ref]
server_name = self._server_name(ref)
LOG.info("removing breakpoint on %s", server_name)
stack_id = next(x['href'] for x in res.links if
x['rel'] == 'stack').rsplit('/', 1)[1]
self.heatclient.resources.signal(
stack_id=stack_id,
resource_name=res.logical_resource_id,
data={'unset_hook': self.hook_type})
succeeds.append(ref)
except Exception as err:
LOG.error("failed to remove breakpoint on %s: %s",
server_name or ref, err)
fails.append(ref)
return (succeeds, fails)
def get_status(self):
self.stack = self.heatclient.stacks.get(self.stack.id)
# check if any of deployments' child resource has last
# event indicating that it has reached a breakpoint (this
# seems to be the only way how to check pre-create breakpoints ATM)
resources = self._resources_by_state()
if self.stack.status == 'IN_PROGRESS':
if resources['on_breakpoint']:
if resources['in_progress']:
status = 'IN_PROGRESS'
else:
status = 'WAITING'
else:
status = 'IN_PROGRESS'
else:
status = self.stack.status
LOG.debug('%s status: %s', self.stack.stack_name, status)
return (status, resources)
def do_interactive_update(self):
status, _ = self.get_status()
# wait for the stack-update to start
while status in ['COMPLETE', 'FAILED']:
status, _ = self.get_status()
time.sleep(5)
while status not in ['COMPLETE', 'FAILED']:
status, resources = self.get_status()
print(status)
if status == 'WAITING':
for state in resources:
if resources[state]:
print("{0}: {1}".format(state, self._server_names(
resources[state].keys())))
user_input = six.moves.input(
"Breakpoint reached, continue? Regexp or "
"Enter=proceed (will clear %s), "
"C-c=quit interactive mode: "
% resources['on_breakpoint'].keys()[-1])
refs = self._input_to_refs(
user_input.strip(),
resources['on_breakpoint'].keys())
self.clear_breakpoints(refs)
time.sleep(5)
print('update finished with status {0}'.format(status))
def _resources_by_state(self):
resources = {
'not_started': {},
'in_progress': {},
'on_breakpoint': {},
'completed': {},
'failed': {},
}
all_resources = self.heatclient.resources.list(
self.stack.id, nested_depth=self.nested_depth)
if self.hook_type == 'pre-create':
hook_reason = 'CREATE paused until Hook pre-create is cleared'
hook_clear_reason = 'Hook pre-create is cleared'
else:
hook_reason = 'UPDATE paused until Hook pre-update is cleared'
hook_clear_reason = 'Hook pre-update is cleared'
stack_change_time = self._stack_change_time()
for res in all_resources:
if self.hook_resource:
if not fnmatch.fnmatchcase(res.resource_name,
self.hook_resource):
continue
stack_name, stack_id = next(
x['href'] for x in res.links if
x['rel'] == 'stack').rsplit('/', 2)[1:]
try:
events = self.heatclient.events.list(
stack_id=stack_id,
resource_name=res.logical_resource_id,
sort_dir='asc')
except heatclient.exc.HTTPNotFound:
events = []
state = 'not_started'
for ev in events:
# ignore events older than start of the last stack change
if ev.event_time < stack_change_time:
continue
if ev.resource_status_reason == hook_reason:
state = 'on_breakpoint'
elif ev.resource_status_reason == hook_clear_reason:
state = 'in_progress'
elif ev.resource_status in ('CREATE_IN_PROGRESS',
'UPDATE_IN_PROGRESS'):
state = 'in_progress'
elif ev.resource_status in ('CREATE_COMPLETE',
'UPDATE_COMPLETE'):
state = 'completed'
resources[state][res.physical_resource_id] = res
return resources
def _stack_change_time(self):
if self.hook_type == 'pre-create':
status_reason = 'Stack CREATE started'
else:
status_reason = 'Stack UPDATE started'
events = self.heatclient.events.list(
stack_id=self.stack.id,
sort_dir='desc')
try:
ev = next(e for e in events if
e.resource_status_reason == status_reason)
return ev.event_time
except StopIteration:
return None
def _server_names(self, deployment_ids):
return [self._server_name(i) for i in deployment_ids]
def _server_name(self, deployment_id):
name = self.server_names.get(deployment_id)
if not name:
if not self.servers:
self.servers = self.novaclient.servers.list()
depl = self.heatclient.software_deployments.get(deployment_id)
name = next(server.name for server in self.servers if
server.id == depl.server_id)
self.server_names[deployment_id] = name
return name
def _input_to_refs(self, regexp, refs):
if regexp:
try:
pattern = "\A{0}\Z".format(regexp)
return [ref for ref in refs if
re.match(pattern, self._server_name(ref))]
except re.error as err:
LOG.warning("'%s' is invalid regular expression: %s",
regexp.encode('string-escape'), err)
return []
else:
return [refs.pop()]