Due to Python's GIL [1], we can't use multiple threads for running
OpenStack services without a performance penalty, since the execution
ends up serialized, which defeats the purpose.
Instead, we should use several processes, since this approach doesn't
have this limitation.
[1] https://wiki.python.org/moin/GlobalInterpreterLock
Change-Id: I09b4dfa0873c5a086d4492a7e2ebb72460e507c9
The RabbitMQ acceptance tests used to be pointing towards the Kilo
repos. This patch addresses it so that it leverages the Puppet
OpenStack integration manifests instead.
Change-Id: I3a647184141f5d66f212ee91d4cf36e8547c7517
Leave the version fields blank, since the release notes document
applies to all versions.
That will avoid manual changes in the future like we did until now.
Change-Id: I00d964aacdb80fe08c85c41a0700e92249d0e2f6
After releasing Pike M3, we need to bump the metadata version
information for the next release with a -dev.
Change-Id: Iffef5bd317125045f6028bcc8c1332abde9323e9
2017-07-20 14:41:23.253791 | manifests/db/mysql/host_access.pp:52:WARNING: arrow should be on the right operand's line
2017-07-20 14:41:23.253911 | manifests/db/mysql/host_access.pp:62:WARNING: arrow should be on the right operand's line
2017-07-20 14:41:23.253949 | manifests/policy/base.pp:43:WARNING: arrow should be on the right operand's line
2017-07-20 14:41:23.253987 | manifests/service_validation.pp:94:WARNING: arrow should be on the right operand's line
Change-Id: I4cee893ae8f2e430149e2cd56b2e9d9d592a75d5
When we switched to the collector as part of
Id09c3358c5843510e6a2a8c0e2d4aeb3607e098b this included using name ==
'httpd' but the package name is different between RHEL/Ubuntu so it
needs to be title == 'httpd' because the title[0] is fixed to be 'httpd'
[0] https://github.com/puppetlabs/puppetlabs-apache/blob/master/manifests/package.pp#L32
Change-Id: I9d535ab38afea852559df2b3073bd4b74a2a3947
The latest version of puppet now reports these as catalog failures so
this change removes the unnecessary references and the references
should be updated.
Change-Id: Id09c3358c5843510e6a2a8c0e2d4aeb3607e098b
Closes-Bug: #1702964
if wsgi_script file is a symlink to another file (eg.
in a virtual environment) the file resource copied the
symlink instead of creating a new file.
This patch ensures, that symlinks are followed and a
ordinary file is created instead of copying the symlink.
Change-Id: I53e59dbfb8810a519fce24e1e381e0f83c5a1c29
As part of the docs migration work[0] for Pike we need to switch to use the
openstackdocstheme.
[0]https://review.openstack.org/#/c/472275/
Change-Id: If4cd490ef9c2ff3f5971f41666f1862b66a1f9ca
If a user has additional code in their openrc file that isn't just bash
variables, the auth provider might fail while parsing it. This change
updates the logic to only try and parse lines with OS_ in it to make
sure that extra things like bash code are ignored.
Change-Id: Id6f5406dcf15642bc0d70caeac30224114bb0669
Closes-Bug: #1699950
Pin reno version to reno!=2.0.0,!=2.3.1,>=1.8.0
Pin sphinx version to sphinx!=1.6.1,>=1.5.1
reno=2.3.1 and sphinx=1.6.1 break releasenotes and doc tests.
Change-Id: I3ac916bbe86cdb9275b2f5c97e747cc2e01eef94
We're coming up on the Pike M2 release and we need to prepare the
metadata.json and reno configuration for the release.
Change-Id: Ic2f64fc0a77bf0f1e3c3776005041c24d5e3dd04
Update the tox configuration to pull in the openstack
upper-constraints.txt when running releasenotes. This will
fix the releasenotes job that is currently failing due to
a new version of sphinx. Additionally this change includes
updates from puppet-modulesync-configs.
Change-Id: I5e38f9490f06e189c60e1ddb390f2eed5e2e7889
After releasing Pike M1, we need to bump the metadata version
information for the next release with a -dev.
Change-Id: Ie00f8f21337242d999a9b034813831ef2aedea80
Currently we only allow to use strings
as the data type for ports.
Due to this we need to cast the data type
in the puppet modules because from THT
this is configured as a Number.
This submission allow to use either string
or numbers for the port parameter in the
os_transport_url function.
Change-Id: I9e56f8e2de542b20fe9e6995506cff5bb435e220
Closes-Bug: #1664561
The ssl paramter of the os_transport_url should be 0/1 but we do not
ensure that this is the value being used in the creation of the
transport_url. This change leverages the stdlib str2bool and bool2num to
ensure that when provided the ssl paramter ends up being 0/1 which is
the correct format for the transport url. Additionally this also allows
for string booleans to be passed into the ssl parameter and it should
still work correctly.
Change-Id: I278559ba98ba48974ae590dc1e0bb99b6ab961d7
