| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The loguserdata.py file gets uploaded to the servers created by Heat to run
under cloud-init. Since the default versions of Python installed on the
user's server may be very old (e.g. RHEL 5 defaults to Python 2.4), avoid
using the octal syntax introduced for Python 3.0 and backported only as far
as Python 2.6. (Also avoid the old syntax, which will break on Python 3.x.)
Also remove use of the "with" statement from loguserdata.py and
part-handler.py. This statement is only available from Python 2.6 on (or in
Python 2.5 via "from __future__ import with_statement").
Finally, remove use of the "except ExceptionType as value" syntax for
catching exceptions. Again, this was only backported to Python 2.6.
Change-Id: I89e86d00993d51e2514b1e589503c6d966909403
Partial-Bug: #1375864
(cherry picked from commit 95ec13c5725ccc19dfe910fe83eb9bc6799d482e)
(cherry picked from commit 140d66e587bce27523ee7ebbeaecbced83082949)
|
|
|
|
| |
Change-Id: I310b093bcc4a6789ca81e1e3ef54b04a55380d7e
|
|
|
|
| |
Change-Id: Iaacffb3164e9353403ac008630c9730fd95422e7
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
With MySQL 5.6 setup to use UTF-8 for everything, the 033 migration
will fail. This is because a VARCHAR(256) cannot be indexed in UTF-8
databases. In versions of MySQL before 5.6, this produced a warning and
a truncated, less-useful index. In 5.6 it was changed to an error.
We will reduce these columns further to 64 bytes in the model in a
subsequent fix. For now, we will just create the table with 64 character
columns on new systems, which will allow this migration to work properly
on mysql-5.6. There is no danger of overflow, as 64 chars is what Keystone
uses for them anyway.
Closes-Bug: #1316646
(cherry picked from commit db0a7fe91d3eddc6d337028332fbd496bb000df2)
Change-Id: I27b9c61828301ba8c392b3d316f656641af753bc
|
| |
| |
| |
| | |
Change-Id: I4fa7090317eb140b3f1e5e1fbc96096b90bcbc95
|
|\ \ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Calling list_join on the list of ResourceGroup's attributes in a
SoftwareConfig fails during template validation with:
ERROR: Property error : my_config: config "list_join" must operate on a
list
Closes-Bug: 1358831
Change-Id: I39d2da7a28ba53fbe3a95bff0198fa5c2f87aa3b
(cherry picked from commit 61a9d167c0f36a2846e39fd33a77ce2755801fe6)
|
|/ /
| |
| |
| | |
Change-Id: I19d4441dda689e3be302bd25b1b59d8a7e5e2904
|
|/
|
|
| |
Change-Id: I712751fc168fa2a34b0ed4219baa7f9376968ce3
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Changes:
- Removed duplicate delete calls to prevent race condition.
- Removed duplicate mocked calls from unit-test.
Closes-Bug: #1311533
Closes-Bug: #1298350
Conflicts:
heat/engine/resources/volume.py
heat/tests/test_volume.py
Change-Id: I5f16c528652f12440160f03b92f41b76d1c9100c
(cherry picked from commit d1ffbd4bfde0cd6b7a82a48b9d4f59cc8b310bd8)
|
|\ |
|
| |
| |
| |
| | |
Change-Id: I679d3f52545af840ee5cbc87a05dae9e54416933
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use the -e parameter in template-validate to override any
Default values specified in the template. This has the desirable
effect of making Horizon use parameters updated from the environment
file.
Conflicts:
heat/engine/service.py
heat/tests/test_validate.py
Change-Id: I84405d1d2f5245ab7cb8b909c2ea1407d5b2973e
Closes-Bug: #1388985
(cherry picked from commit 09bc7ad79e4d95cffd8db19c781dc66230640d84)
(cherry picked from commit dc1b99d064d9fbb6439357e89c770f15810f7cce)
|
|
|
|
|
|
|
|
|
|
|
|
| |
There is a situation, when we request resource data
from db during loadbalancer resource deleting. If
there is no such data, NotFound exception raised and
stack delete failed with this error. To prevent this
behaviour, there is a need to catch NotFound exception
in delete method of loadbalancer.
Change-Id: Ie52bc3059e91bd25effc53bbd255bce6948509ac
Closes-bug: #1381700
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Check the domain ID for the project matches the configured domain from
heat.conf, so we avoid the bad outcome of deleting the user's project
when heat has been configured for stack domain users after creating
some stacks.
Conflicts:
heat/common/heat_keystoneclient.py
heat/tests/test_heatclient.py
Change-Id: I247e732033f44daf0de5b0efeff6c263814d13ab
Closes-Bug: #1365332
|
|\ \
| | |
| | |
| | | |
stable/icehouse
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We have PERSONALITY property on OS::Nova::Server that defaults to {},
but code that triggers check for limits compares value of property
to None -> this property always get validated for limits (if limit is
set).
Change-Id: I552a03cd9f4ab570aa059349ab361bb50c9c0d1e
Closes-bug: #1333283
(cherry picked from commit 78b68ba0da5d6ad71397538e484fb7565333129b)
|
| | |
| | |
| | |
| | | |
Change-Id: I1b2e9b58d2eccbe4e7a00b1549adcdcea5845afb
|
|\ \ \ |
|
| | |/
| |/|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If delete a alarm which has been deleted from ceilometer
already, ceilometer client will throw a exception, and deletion
will fail, so we need to ignore NotFoundException when
deleting alarm.
Change-Id: I224c2de575106d3ce827d48853e0729b6bf9a221
(cherry picked from commit c661482b2edace9f8bded182548ba0db7a2cdafa)
|
| |/
|/|
| |
| | |
Change-Id: I6838e089c1830fd1cc318f10910381b4d59a2ed4
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The software_deployment resources lazy loads data via the
heatclient _get() function. Not only is this function private
but the oslo-incubator recently reverted things such that
get() should once again be used for lazy loading details.
See oslo-incubator commit 04a1abe59ac39890f57d2eed57d3d1b12bb5d757
This was merged into heatclient master in:
72017c566850da589f7dc7dc275871986aadbd54
This patch makes it so that heat will work with both versions
of heatclient.
Change-Id: I1799636d39cea410b55ffe72eeecba4f4a7a0f80
Closes-bug: #1308770
(cherry picked from commit a7d782395084c6c16304f17d14b824028c5ba450)
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
New tox (>=1.7.0) sets a random python hash seed by default. This is
generally good for testing because it will help keep projects working
regardless of the hash seed, but heat unittests don't currently pass
with a random hash seed so set it to the python default seed.
This change will allow us to use new tox again and remove the
restriction on tox<=1.6.1 to run unittests.
Redundant setenvs for specific test envs are removed to prevent them
from overriding the PYTHONHASHSEED value.
Note this change will need to be backported to the stable branches to
keep unittests there working with new tox as well.
Change-Id: I2370759546adee7bf5e20fa214ed7fa1f31d6bfe
(cherry picked from commit 47e4a7e12f2287cd938c652f56e96a108d59bfd5)
|
| |
| |
| |
| |
| |
| |
| | |
Bump stable/icehouse next version to 2014.1.3
Change-Id: I9256ef90cc4546cccd5c5d6af6432e6ca3e3a99e
Signed-off-by: Chuck Short <chuck.short@canonical.com>
|
|\ \ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
In the case where a stack that has a watch task associated with it is
being deleted and the engine successfully acquires a lock, no attempt to
stop the stack will be made since it has no action in-progress. This
will result in orphaned watcher tasks being left behind, since the
watcher tasks are normally stopped as part of the call to
ThreadGroup.stop().
To fix this, explicitly stop the periodic watcher tasks before calling
stack.delete().
Conflicts:
heat/engine/service.py
Change-Id: Idf5b7c1100fc85743ac6dc5831731f7ecc96bfd6
Closes-Bug: #1315044
(cherry picked from commit 33ea3573422ac32f4bd1788603fef64bea7e85bd)
|
|\ \ \
| |/ / |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This is to allow change Idf5b7c1100fc85743ac6dc5831731f7ecc96bfd6 to be
applied to the stable branch.
Change-Id: I839129874de6697594fbe5f3bfb937180041bdbc
|
|\ \ \
| |/ / |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Applying novaclient exceptions fix from commit:
9850832da563c1113e121b99d38feb21af9daa8d
"Raising nova exceptions has been moved to
heat.tests.v1_1.fakes.fake_exception since it is actually impossible
to directly create a nova exception object which works on both
novaclient 2.17 and (yet to be released) 2.18.
nova.exceptions.from_response is stable across novaclient releases
so all test exceptions are now raised using from_response."
Change-Id: I0c5e58ffdd3eb13b9a87df22f1c5e06cec972913
Closes-Bug: #1322098
|
|\ \ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We make sure that there is no infinite recursion by not repeating
a TemplateResource within a resources inheritance.
Change-Id: I1ffa66886923b15b52a8075089dd4042f0a55862
Closes-Bug: #1297396
|
|\ \ \ \
| |_|_|/
|/| | | |
|
| | |/
| |/|
| | |
| | | |
Change-Id: Ib84db5aaf53e8f44efde0ae400ac6fa4187f27df
|
|\ \ \
| |/ /
|/| | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If the metadata of a provider resource changes, we shouldn't replace the
whole provider.
This is not a backport, because this has been fixed more generically for
all resources in master, where update_allowed_properties no longer
exists.
Change-Id: I706b3a2faeef582e906fe2195d79200ab3853a2c
Closes-Bug: #1332354
|
|\ \ \
| |_|/
|/| | |
|
| |/
| |
| |
| |
| |
| |
| |
| |
| | |
Add unicode encode to parameter to make sure unicode will be
accepted correctly.
Change-Id: Idc9afad7971eaf0320d719946a3f09b8e15c1479
Closes-Bug: #1330370
(cherry picked from commit 7b8d32d58195a2b9098c837fe997d334e2529b8d)
|
|/
|
|
| |
Change-Id: Id7181755d3609c0ae91c2402b34f2de02c06a1f0
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch disables validation of stack parameters values when
loading existing stacks from the database. The old behavior was
to always validate stack parameter values when initializing a
Stack object. However, this causes problems especially with
custom constraints when a constraint could be fulfilled during
stack creation but cannot be fulfilled at a later point (e.g.
when a flavor or keypair has been deleted). In such a case,
the existing stack became completely unusable.
This patch changes the behavior to not do parameter validation
for existing stacks. The assumption is that validation at stack
creation time should be sufficient to make sure the stack is valid,
so only valid stacks end up in the database. Therefore, validating
parameters again when loading stacks is not really required.
Closes-Bug: #1314401
Cherry-pick from review https://review.openstack.org/#/c/94329
(cherry picked from commit 349a6f04b53309f7fa8f3225f98d41670888d20c)
Conflicts:
heat/tests/test_parser.py
Change-Id: I0947c2dcfb9db4c81b07de582aab86f262d0c008
|
|\ \
| |/ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This patch moves validation of constraints out of the constructor of
schema objects into a separate method that can be invoked on the
schema objects after creation. This gives us more control over when
validation shall be invoked.
This patch fixes an issue with custom constraints that could not be
validated when having a default value, since they require the RPC
context to be present. However, the context was not present in the
places where needed.
The short-sighted fix would have been to change a lot of method
signatures to pass the context properly, but that did not seem to be
really clean. The better fix seemed to be to move validation into
a separate method that can be invoked with the proper context. This
will also give us more control on when we do validation, and which
validation we do.
This patch
Closes-Bug: #1314240
This patch will also enable to fix
Partial-Bug: #1314401
For that bug it will be necessary to switch off validation under
certain conditions, which was not possible before when doing
validation right in the constructor.
Cherry-pick from review https://review.openstack.org/#/c/91485
(cherry picked from commit 19f83e84269e1aec2d299ada616a4648e00f4333)
Conflicts:
heat/engine/parameters.py
heat/engine/service.py
heat/tests/test_hot.py
Change-Id: I19e1fb520551eb42bf36bb380f1cc28c81bbaedd
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
| |
This change is already merged in oslo-incubator. Original commit
message body includes:
Porting this fix from oslo.messaging. This fixes the impl_qpid.py
driver to allow it to work with the latest stable upstream QPID broker
(version 0.28). See the Apache Qpid Jira bug
https://issues.apache.org/jira/browse/QPID-5557
Change-Id: If71f78e50f8a9b3acfd1e9d02c8271f17c4ebee7
Related-Bug: #1300318
(cherry picked from commit 3597ffca77133101943114b80fd4263bada87f4f)
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
When a signal is received and processed by a Resource, there could be an
event generated automatically. For such an event, Heat is supposed to
provide some useful message (including status_reason). But currently,
the resource code can only process ceilometer alarms and watch rules,
for signals sent for a SoftwareDeployment resource, the signal is only
providing an string 'Unknown' as the status change reason. This is
confusing when a user performs 'heat event-list stack'.
This patch adds an additional check (guess) whether the signal comes
from a SoftwareDeployment. If that is the case, fill in the status
reason as 'deployment succeeded', or 'deployment failed (n)', where n is
the deploy_status_code that comes with the signal.
Closes-Bug: #1300679
Change-Id: Ic970a5dac923fffdcdd365983df58f4d338de849
|
|\ \ |
|