Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Removing the master and slave vocabulary #1

Open
wants to merge 966 commits into
base: master
Choose a base branch
from

Conversation

bopopescu
Copy link

For diversity reasons, it would be nice to try to avoid 'master' and 'slave' terminology in this repository which can be associated to slavery. The master-slave terminology could be problematic for people in several countries which has the history of slavery like Romania, USA and many others. Thank you for considering the proposal. Let me know if any changes in the PR are needed, I would be happy to implement them.

mdbooth and others added 30 commits August 15, 2014 16:07
spawn() contained an important boolean called ebs_root. This was True
for any spawn which contained a block device mapping. This no longer
has any meaning, and was confusing. This change removes references to
ebs_root, instead using block_device_info directly.

N.B. The block device mapping behaviour of the VMware driver is still
not correct, but fixing it is not the purpose of this patch. This
patch simply makes the existing behaviour easier to see.

N.B. test_spawn_mask_block_device_info_password in test_vmops.py
assumes internal details of spawn in order to cause a failure after
emitting a log message. In moving the log message we need to update
the dance this test does to ensure that spawn fails after the log
message, but not before.

This change has been split out of
https://review.openstack.org/#/c/87002/, which was written by Shawn
Hartsock.

partial blueprint vmware-spawn-refactor

Co-authored-by: Shawn Hartsock <[email protected]>
Change-Id: I11b355a6b53194bc9f85f542b54c9fbb9c061e04
FakeVim._create_vm() was creating a VM with a default datastore rather
than the given one. This prevented testing that the correct datastore
has been used.

Change-Id: I640a21f23b2d5c1186772c8c5ce3ca8927e6cbc5
config.hardware.device was handled inconsistently, and could be of
various types. When created by _create_vm(), it was actually a
VirtualDeviceConfigSpec[]. It should be a VirtualDevice[].

Fix tests which depended on the wrong type.

Add _create_array_of_type function to fake to create an array type
which works correctly when passed to code. A returned array of Foo
must be an instance of class ArrayOfFoo, and contain a single
property, Foo, which is a python array of objects of type Foo.

Change-Id: Ic90e1d0d947941621606fded9e56f9a98d111cad
Move the spawn() code which builds a new VM into
vmops.build_virtual_machine().

We remove the duplicate check of instance_name. Other than that this
is straight code motion.

In addition, we add a basic test of the function of
build_virtual_machine, which tests that all given configuration is
reflected in the resulting virtual machine.

This change has been split out of
https://review.openstack.org/#/c/87002/, which was written by Shawn
Hartsock.

partial blueprint vmware-spawn-refactor

Co-authored-by: Shawn Hartsock <[email protected]>
Change-Id: Id9c95966eefb644fdaf36465e7b3d3583826314a
Factor out the code to configure the virtual machine with a new
config drive iso image.

partial blueprint vmware-spawn-refactor
Change-Id: Ie387445ae83e89fc436243a4c8ee9030d5221510
Update the tests to ensure that the expected parameter is first.

TrivialFix

Change-Id: I5a860d7a217f69bc1d5e7e0e470e724e62a69372
The class was only used by the ESX driver which was remove in the
commit 1deb31f.

Change-Id: I026e4c0f463e78e33592c6408d2e688c74a0714b
This support is no longer needed as we only support the VC driver
since commit 1deb31f.

Change-Id: Iad71fceb580620b1843ce3cf5d926343d8113945
Separates the unit tests of the file transfer download module (the only
one that is implemented in nova.image.download) out into its own unit
test file test_transfer_modules.py out of test_glance.py. In the
process, properly limited the unit test boundaries of the transfer module
unit tests to just the transfer module class' download() method, instead
of testing all the way through the GlanceImageService class
unnecessarily (since that is already tested thoroughly in
test_glance.py). As an added benefit, the unit tests of the download and
transfer module stuff went from around 4 seconds to less than a
quarter-second for all the tests.

Change-Id: I2ff6e31359d343b78dc4d03cf2beafd8a99b0a29
Refactors the tests in nova.tests.image.test_glance that were checking
the behaviour of both the GlanceClientWrapper retry logic and the
glanceclient.Client creation to use mock instead of the FakeGlanceClient
in nova.tests.glance.stubs. The fake stub client was actually masking
issues in the existing test cases, including not properly checking the
identity headers that are actually supplied to the real
glanceclient.Client constructor.

Change-Id: I1c114df8e4ab2fccd966ed4af22181881590c443
Partial-bug: #1293938
Removes the final piece of glance_stubs from the image unit tests.

Change-Id: I0db3b6c83edaf91466e85d423ce75b3e75fd3517
Closes-bug: #1293938
This adds another developer "helper trap" for inter-object version
dependencies. If an object that another object depends on is bumped
in version, this test will note the change and remind the developer
(or reviewer) that the parent needs a version bump and backport
rule added.

Change-Id: Ia81d034f850173401cd84a79b15ae0f7c80b9653
The logic to identify volumes is currently a nested function
in _default_block_device_names, named _is_mapping. It should
be moved to a more general location so others could utilize it
and allow it to be properly unit tested. The following patch
moves _is_mapping to nova/virt/block_device.py and renames it
to is_block_device_mapping.

Change-Id: I560abc4b57ca5bd195282af7cd1ab9bbf7600b67
Closes-Bug: #1351810
Jenkins and others added 30 commits August 26, 2014 09:52
Commit da66d50 restricted
attaching external networks to admin clients. This patch changes
it to a policy based check instead with the default setting being
admin only. This allows operators to more precisely configure who
they wish to allow to attach external networks without having to
give them admin access

Change-Id: I59e71f117f889f2abffddc36c1870ef1e0fe3711
DocImpact: Adds network:attach_external_network policy
Closes-Bug: #1352102
Adds a filter AggregateDiskFilter which provides the ability to
read from aggregates metadata the "disk_allocation_ratio".

DocImpact
Implements: blueprint per-aggregate-disk-allocation-ratio
Change-Id: I79d59fcdfb09e67ed6f12113615c673624b24a19
The service list version information does not match, fix it.

Change-Id: Ibc88aedcf63bbcc5a36fd3a3bd92d0a0f0424fb2
This is based on a new QEMU feature called vhost-user for connecting
to a user-space vswitch via a unix socket.

Change-Id: I0a11e89d3a0bbdfccac569c09009e576ea2e1173
Signed-off-by: Michele Paolino <[email protected]>
… < 1.3.0

Known moderate severity security vulnerability detected in SQLAlchemy < 1.3.0
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.