openstack
/
heat
Code Issues Proposed changes

Respect locks when changing FAILED resource to COMPLETE 

If a resource is in a FAILED state, but the plugin overrides
_needs_update() to not raise UpdateReplace, and also no actual update is
required, we simply change the resource's state to UPDATE_COMPLETE.
However, since the locking/unlocking steps were merged with the resource
state updates in a7376f7494, in doing so we
were ignoring the lock (since at this stage the resource is still unlocked,
and we don't need to take a lock). If another update traversal had acquired
the lock in the meantime, we would still write to the locked resource.

This change adds a new LOCK_RESPECT lock mode to Resource.store(). In this
mode the lock is not acquired or released, but an error will be raised if
the resource is found to be locked.

Change-Id: I8b5cd1e05b88dd13abc13899a73f23810f7e6135
Closes-Bug: #1705794
Related-Bug: #1662585
This commit is contained in:
Zane Bitter 2017-07-21 21:32:45 -04:00
parent bdddeee602
commit 1da0790e0d
1 changed files with 14 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
heat/engine/resource.py
View File

@ -98,8 +98,11 @@ class PollDelay(Exception):
class Resource(status.ResourceStatus):
BASE_ATTRIBUTES = (SHOW, ) = (attributes.SHOW_ATTR, )
LOCK_ACTIONS = (LOCK_NONE, LOCK_ACQUIRE, LOCK_RELEASE) = (
0, 1, -1)
LOCK_ACTIONS = (
LOCK_NONE, LOCK_ACQUIRE, LOCK_RELEASE, LOCK_RESPECT,
) = (
None, 1, -1, 0,
)
# If True, this resource must be created before it can be referenced.
strict_dependency = True
@ -1297,22 +1300,7 @@ class Resource(status.ResourceStatus):
if not persist:
return
rs = {'current_template_id': self.current_template_id,
'updated_at': self.updated_time,
'requires': self.requires}
if not resource_objects.Resource.select_and_update_by_id(
self.context, self.id, rs, expected_engine_id=None,
atomic_key=self._atomic_key):
LOG.info("Resource %s is locked, can't set template",
six.text_type(self))
LOG.debug('Resource id:%(resource_id)s locked. '
'Expected atomic_key:%(atomic_key)s, '
'accessing from engine_id:%(engine_id)s',
{'resource_id': self.id,
'atomic_key': self._atomic_key,
'engine_id': self._calling_engine_id})
raise exception.UpdateInProgress(self.name)
self._incr_atomic_key(self._atomic_key)
self.store(lock=self.LOCK_RESPECT)
self._calling_engine_id = engine_id
registry = new_stack.env.registry
@ -1502,8 +1490,10 @@ class Resource(status.ResourceStatus):
status_reason = _('Update status to COMPLETE for '
'FAILED resource neither update '
'nor replace.')
lock = (self.LOCK_RESPECT if self.stack.convergence
else self.LOCK_NONE)
self.state_set(self.action, self.COMPLETE,
status_reason)
status_reason, lock=lock)
return
if not self.stack.convergence:
@ -1969,9 +1959,13 @@ class Resource(status.ResourceStatus):
if lock == self.LOCK_ACQUIRE:
rs['engine_id'] = self._calling_engine_id
expected_engine_id = None
else: # self.LOCK_RELEASE
elif lock == self.LOCK_RESPECT:
expected_engine_id = None
elif lock == self.LOCK_RELEASE:
expected_engine_id = self._calling_engine_id
rs['engine_id'] = None
else:
assert False, "Invalid lock action: %s" % lock
if resource_objects.Resource.select_and_update_by_id(
self.context, self.id, rs, expected_engine_id,
self._atomic_key):