Age | Commit message (Collapse) | Author | Files | Lines |
|
Added support for obtaining an initialized
OpenStackHeatStack object from an existing Stack object
JIRA: SNAPS-278
Change-Id: I68cb0c85179868e4e3899b5516316dda3321e9f6
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
JIRA: SNAPS-286
Change-Id: Ib4976529168655df0ed906276b503b70a8459f21
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
JIRA: SNAPS-215
Change-Id: I94923a9f184b0d370159b499919e8fc20a2fef90
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
By not closing all of the keystone sessions being created when
running all of the tests, this may be the root cause to the IOError
occasionally being observed:
IOError: [Errno 24] Too many open files
JIRA: SNAPS-285
Change-Id: I7fc7ab0c6cdd02f1ae32bb3ae4f121cb465d5693
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
and compute_host names while deprecating the method get_vm_info().
JIRA: SNAPS-277
Change-Id: Idc8578b3f2cf2be8ef90f52dd025dbea729b222b
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Although this patch has added the project_name member to
VolumeConfig, it does not work properly due to a cinder bug
https://bugs.launchpad.net/cinder/+bug/1641982. However, the
code will now allow for volumes with the same name to be
created on the same pod just as long as they do not exist
within the same project/tenant.
JIRA: SNAPS-270
Change-Id: Ia6f98e9b66793a69c0205d2d8e38bb3d03f3333e
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Changed super test class OSIntegrationTestCase to allow subclasses
to set the user roles when required. For all heat stack test,
the user role will always be 'heat_stack_owner'. 'admin' will be
used for tests requiring this role (i.e. heat template creates
flavors)
JIRA: SNAPS-268
Change-Id: I0ec01cfea6297547a6a158eec8e5c69af6d6c7d7
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Routers should be able to be created by admin users to other
projects. Routers without admin users should only be able to
create ones to their default project.
JIRA: SNAPS-266
Change-Id: I5ecfbcc2ed94bbab211751da9521f3ec1d4ff66d
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
JIRA: SNAPS-276
Change-Id: Ia676b91ee881097ca7502fb4b9f5c64961de2214
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
The pattern being replaced has unwittingly added the requirement that
all creator credentials must be of type 'admin' as when looking up
the associated project ID required a call to keystone.projects.list().
As the SNAPS integration tests were always creating users with an 'admin'
role, this issue was not caught. As part of this patch, integration test
users will no longer be admin.
JIRA: SNAPS-274
Change-Id: I02957f69e31a9d4dfa63362d371f061687e59fbf
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
The 'test_add_remove_volum' is supposed to wait for volume attachment
and detachments. However, the code for the latter was missing. Moreover,
we remove some code that checks for attached volumes since that should
never happen when detaching the volume.
Change-Id: I9ed54a467062d0ddc4efeddf03af7ebe6173e789
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
Patch I2245adb84d0e6d4e3350d17d0c2e44baf5202d51 increased the timeout
for the API calls to 2 minutes which was then somewhat reverted in
Iea3aeab59c378917fbd175d673113e8d30e2e4b9. However, the problem still
persists on slow systems and 30 seconds may not be enough to confirm
a volume attachment to the VM.
Change-Id: I860b2a797917e13772e95c5d756919e9792b63a6
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
JIRA: SNAPS-265
Change-Id: If58f5817e52e21be377d2b06a5f7c3a413454350
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
security groups with the same name from a different project
JIRA: SNAPS-264
Change-Id: I7c905e5588d5e503b47ea0e9f5997be89e841aec
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
|
|
It could have worked thanks to interned strings.
Change-Id: If56cc44fe35ca06bc9033af620ddd3c0586877fe
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
Change-Id: Ib226ef99eed2fda2c7112faf0f0e9339afed90fe
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Change-Id: Idb6912319e24edb39e18114bd3c1ef5bca750752
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
It completes the previous patch [1] about that topic.
[1] https://gerrit.opnfv.org/gerrit/#/c/51845/
Change-Id: Ib2ba07c0d5608bd452ed209544d4dbfc17124299
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
That could easily raise side effect as it overrides default configs
set by Cloud providers.
Change-Id: I659d16f8f58bba3a07b8ae2ee2270aa69c200ed5
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
The timeout logic in nova_utils.attach_volume() and detach_volume()
was not correct which may have been the root cause behind the issue
FUNCTEST-927. Timeout in both attach and detach is no longer
optional.
Also added a test to attach and detach without timeout as that path
was never tested. Updated associated test documentation as well.
JIRA: SNAPS-263
JIRA: FUNCTEST-927
Change-Id: Iea3aeab59c378917fbd175d673113e8d30e2e4b9
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
Change-Id: Idd39e4f6e3f22fcbafe3aae2512ac36cf9e2a417
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
The optional parameter 'project_name' was never getting
a default value when not used causing an OSCredsError
to be raised.
Change-Id: I2a0986c62c5d9520f60ebd06db1f3a749e62361c
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
On slow environments, time.sleep(10) is not enough for attach/detach API
calls to propage properly leading to failures like the following one
(reduced api_check test)
2018-01-31 18:53:19,297 - functest.ci.run_tests - INFO - Running test case 'api_check'...
2018-01-31 18:54:27,434 - functest.ci.run_tests - INFO - Test result:
+-------------------+------------------+------------------+----------------+
| TEST CASE | PROJECT | DURATION | RESULT |
+-------------------+------------------+------------------+----------------+
| api_check | functest | 01:07 | FAIL |
+-------------------+------------------+------------------+----------------+
We can workaround this by using the nova_utils.{de,at}tach_volume}
'timeout' argument to wait up to 2 minutes for the attach/detach call to
complete. This doesn't affect normal exception whilst it helps get the
test going on slow systems. WIth this change, the api_check (reduced
test) passes as follows:
2018-01-31 18:48:35,469 - functest.ci.run_tests - INFO - Running test case 'api_check'...
2018-01-31 18:50:56,705 - functest.ci.run_tests - INFO - Test result:
+-------------------+------------------+------------------+----------------+
| TEST CASE | PROJECT | DURATION | RESULT |
+-------------------+------------------+------------------+----------------+
| api_check | functest | 02:20 | PASS |
+-------------------+------------------+------------------+----------------+
Change-Id: I2245adb84d0e6d4e3350d17d0c2e44baf5202d51
Signed-off-by: Markos Chandras <mchandras@suse.de>
|
|
Else OSCreds.__str__() simply doesn't work.
Change-Id: I0a0147862ad3015dc9cf284c61429f1534a47d61
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
configurations.
Change-Id: I35374ea5fa79974c43c2a474e43ff275b26dfc02
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
previous means was not compatible with python-novaclient==10.0.0"
|
|
Change-Id: I227433e92fc7f42924fa9404ff3c394c79500a1b
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
means was not compatible with python-novaclient==10.0.0
JIRA: SNAPS-258
Change-Id: I56f7a8341c81e0ae5596f67d4d15ef6c26e0e680
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Introducing a version check to avoid checking keystone domains,
which are an auth API v3 feature, when only auth API v2.0 is
being used.
Change-Id: I220c6637e5f65124e5dbe7d5246490c3986b7a66
Signed-off-by: Georg Kunz <georg.kunz@ericsson.com>
|
|
Change-Id: Iadb5e44836258e5d9a89090c4b0302d2689a40c5
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
when initializing network resources Enhanced playbook runner variable support"
|
|
Additional protections when initializing network resources
Enhanced playbook runner variable support
Change-Id: Id897c4ece1de706afbc52d0a034ca3bfb68fdce3
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
Change-Id: I120faaa8695db5f83e39b521aaa6c1ab68e7fcc1
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
The only task currently supported is reboot.
Change-Id: I9089c593af30db071508a47667381238fe91e44e
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
* Simplified the configuration of file resources
* Added integration test
JIRA: SNAPS-255
Change-Id: I1065d9352865d7a9f946a5d9947e32e7340f20bc
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
Change-Id: Ib66607b7656093c5f0cd0e2cb4c0aa9c3b9c661a
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
|
|
Added cloud_init_timeout to VM instance configuration to override the
default timeout value hardcoded @ 120 seconds. This patch not only
expands the default value to 300, but also allows the client to override
this value as cloud-init can take quite some time depending on what is
being requested.
Change-Id: Ia4d0a54ce853a93748b16945f5ddd4d5887feb7d
Signed-off-by: spisarski <s.pisarski@cablelabs.com>
|
|
|
|
When the network_type is vlan, the params physical_network and
segmentation_id should be accepted as arguments when creating
network.
JIRA: SNAPS-252
Change-Id: I3a7bec8a592ede65b97eff0911d09a8154349385
Signed-off-by: Linda Wang <wangwulin@huawei.com>
|