summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* | | | | | | | Merge "Use bare container format by default"Jenkins2012-09-051-2/+1
|\ \ \ \ \ \ \ \
| * | | | | | | | Use bare container format by defaultBoris Filippov2012-09-051-2/+1
| | |_|_|/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Set container_format to bare during libvirt snapshot, when VM image in glance was deleted. Currently, if VM image in glance was already deleted before snapshot, nova will attempt to create snapshot image with container_format: None. This cause glance to return error on attempt to upload snapshot. According to glance docs container_format is not used anywhere in glance or nova explicitly and it is safe to set it to bare, when you are unsure which container_format you need to use. Current snapshot logic sets snapshot disk_format to currently used image_format in absence of base image in glance. This resolves bug 921774 without need for snapshot mechanism redesign. Change-Id: I7beea35120aaeac0837daecdf58f38f62e24454c
* | | | | | | | Merge "Sync some updates from openstack-common"Jenkins2012-09-054-6/+6
|\ \ \ \ \ \ \ \ | |_|_|/ / / / / |/| | | | | | |
| * | | | | | | Sync some updates from openstack-commonMark McLoughlin2012-09-054-6/+6
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Syncs the latest code from the stable/folsom branch. Includes: 5a56539 Allow non-string items in the creds dict. 769ec65 Don't trap then re-raise ImportError. 202b8b7 Fix spelling typos Change-Id: I6473358c743ce77b90dfb1282f3e257e6fd3dbf6
* | | | | | | Merge "Yield to another greenthread when some time-consuming task finished."Jenkins2012-09-052-1/+9
|\ \ \ \ \ \ \ | |/ / / / / / |/| | | | | |
| * | | | | | Yield to another greenthread when some time-consuming task finished.TianTian Gao2012-09-052-1/+9
| | |_|_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Partially addresses bug #1045152 On a heavily loaded compute node, it can be observed that periodic tasks take so long to run that the report_state() looping call can be blocked from running long enough that the scheduler thinks the host is dead. Reduce the chance of this happening by yielding to another greenthread after each periodic task has completed and each loop in some methods that has linear relationship with the number of instances. Change-Id: If2b125708da8298b20497e2e08e52280c102f1e1
* | | | | | Merge "Allow hard reboot of a soft rebooting instance"Jenkins2012-09-052-2/+45
|\ \ \ \ \ \
| * | | | | | Allow hard reboot of a soft rebooting instanceMathew Odden2012-09-012-2/+45
| | |_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow hard reboot API calls to succeed if the task status is soft reboot, but not hard reboot. This was discussed in the previous change for this bug in Iae95b121c5c623e1eaa6fba52b13170413722d2d Fixes bug #944015 Change-Id: I0f78faad524774dbbcad0bf2adafed3802dfeb75
* | | | | | Merge "Makes sure new flavors default to is_public=True"Jenkins2012-09-052-1/+40
|\ \ \ \ \ \
| * | | | | | Makes sure new flavors default to is_public=TrueVishvananda Ishaya2012-09-042-1/+40
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The default for is_public is true in the database, but new flavors created through the extension was defaulting is_public=None and passing it through to the db layer where it gets converted to False. This patch fixes the issue by defaulting is_public to True at the api layer. Incldes a broken test to verify. Fixes bug 1046040 Change-Id: I4c361c0055a14de29e364868074e17cf3ec23220
* | | | | | | Merge "Optimizes flavor_access to not make a db request"Jenkins2012-09-051-18/+6
|\| | | | | |
| * | | | | | Optimizes flavor_access to not make a db requestVishvananda Ishaya2012-09-041-18/+6
| | |/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We are now caching the flavor object in the request so we don't need to make another db request. Note that there is further refactoring possible but I wanted to keep the changes small to minimize chance for regression. Also, no need to check for not found since we will have errored much earlier if the flavor doesn't exist. This change is necessary to create a failing test for bug 1046040 Change-Id: I718203b33a192575288057dd4631a9aa3086b567
* / | | | | Automatically convert device namesVishvananda Ishaya2012-09-047-35/+69
|/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the past, users have been able to specify xvda or xvdb and it has worked. We are now validating device names against the expected name from the backend, and this causes confusion, especially for lxc which expecting device names to be /dev/a /dev/b /dev/c etc. This patch addresses the issue by automatically converting between different formats. The proper format for the backend will be returned by the api. Includes tests to verify that the conversion works and that the lxc values work as expected. Fixes bug 1046020 Change-Id: Iffa552ba05f89f70b6fb93043edf8882c8412215
* | | | | Merge "Skip tests in OSX due to readlink compat"Jenkins2012-09-041-0/+5
|\ \ \ \ \
| * | | | | Skip tests in OSX due to readlink compatBrian Lamar2012-09-041-0/+5
| |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Although greadlink can be installed on OSX, there is not compatibility in the code so the tests should be skipped until OpenStack supports running this code on OSX/BSD. Fixes bug 1037835 Change-Id: Iddaba950f9f8b3823bd62b59b36e08aff4a979f8
* / / / / On rebuild, the compute.instance.existsRick Harris2012-09-015-19/+34
|/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | notification should contain the original image metadata. In order to do this, the original system metadata associated with the instance is preserved and passed along to the rebuild_instance() in the compute manager. (Code originally from Iccha Sethi and Rick Harris.) bug 1044008 Change-Id: Ieb60a2871a2a209ef467e2b51dab375872649a18
* | | | Merge "Adds API sample test for limits"Jenkins2012-09-045-0/+213
|\ \ \ \ | | |/ / | |/| |
| * | | Adds API sample test for limitsTiago Mello2012-09-035-0/+213
| | | | | | | | | | | | | | | | | | | | | | | | Adds samples for GET /limits Change-Id: Ie6b42119a9d60273d5d4d0c17fb44082009bc847
* | | | Merge "Add API samples to images api"Jenkins2012-09-0457-3/+2198
|\| | |
| * | | Add API samples to images apiMauro S. M. Rodrigues2012-09-0357-3/+2198
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes issue with updated metata not being returned properly from the api. Adds a small check to the samples baseclass to allow the output to be generated even if the template file is empty. Change-Id: Idcdd24d7e785026e50b9abf4762466fd55a8d42f
* | | | Merge "Add method to manage 'put' requests in api-sample tests"Jenkins2012-09-041-2/+5
|\| | |
| * | | Add method to manage 'put' requests in api-sample testsMauro S. M. Rodrigues2012-09-031-2/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Include a _do_put() method into tests_api_samples to generate api-samples to put requests. Change-Id: I1407f908592fcda8d83506ccb98f17f3af84c13c
* | | | Merge "Fix hyperv driver spawn() signature"Jenkins2012-09-042-4/+6
|\ \ \ \
| * | | | Fix hyperv driver spawn() signatureMark McLoughlin2012-09-032-4/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #1045392 commit 99fb9d2a changed the signature of the spawn() method but not all drivers were updated for the change. The injected_files and admin_password instance attributes were ignored by the hyperv driver before, so we can just ignore them in spawn() now that they are being passed as function arguments. Change-Id: I129d21316488a18a4f062252ba5f788bc65c7c25
* | | | | Merge "Fix vmwareapi driver spawn() signature"Jenkins2012-09-042-3/+5
|\ \ \ \ \
| * | | | | Fix vmwareapi driver spawn() signatureMark McLoughlin2012-09-032-3/+5
| |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #1045393 commit 99fb9d2a changed the signature of the spawn() method but not all drivers were updated for the change. The injected_files and admin_password instance attributes were ignored by the hyperv driver before, so we can just ignore them in spawn() now that they are being passed as function arguments. Change-Id: Ifc4279a05127dcc5c01388882c40570bf6de5ff2
* | | | | Add full python path to test stubbing modules for libvirtIan Wells2012-09-032-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | libvirt test modules were relatively addressed (import fakelibvirt etc.) and this meant that running the test specifically (./run_tests test_virt_drivers) would fail with missing imports. Change-Id: I81f019bf635e10ca90f77f7a30b2e19065fee8f4
* | | | | Merge "Fixes spawn method signature for PowerVM driver."Jenkins2012-09-032-17/+4
|\ \ \ \ \
| * | | | | Fixes spawn method signature for PowerVM driver.Tiago Mello2012-08-292-17/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ComputeDriver has now 2 new required arguments which are: injected_files and admin_password. This patch adds the arguments in the PowerVMDriver spawn signature. Fixes bug 1040770 Change-Id: I0a45fa45c942f2d0381c5a7310d386ae655073fe
* | | | | | Merge "Rename imagebackend arguments."Jenkins2012-09-035-38/+37
|\ \ \ \ \ \
| * | | | | | Rename imagebackend arguments.Michael Still2012-09-035-38/+37
| | |/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rename function arguments to something which more clearly indicates their purpose. No functional changes. Change-Id: Iedbc5477ac54527b60c520fd7c774e608adb07c0
* | | | | | Merge "Fixes sqlalchemy.api.compute_node_get_by_host"Jenkins2012-09-031-1/+1
|\ \ \ \ \ \
| * | | | | | Fixes sqlalchemy.api.compute_node_get_by_hostJian Wen2012-09-031-1/+1
| |/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use join('service') instead of options(joinedload('service')) to get the right compute_node Fixes bug 1044238 Change-Id: I082a46847e63a1e21083a625c3b07ec9be05412f
* | | | | | Merge "Update nova-rpc-zmq-receiver to load nova.conf."Jenkins2012-09-031-8/+9
|\ \ \ \ \ \ | |/ / / / / |/| | | | |
| * | | | | Update nova-rpc-zmq-receiver to load nova.conf.Dan Prince2012-08-301-8/+9
| | |_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Updates the bin/nova-rpc-zmq-receiver so that it makes use of FLAGS, calls parse_args, and loads the nova.conf config file like all other nova- binaries. Fixes LP Bug #1044153 Change-Id: I904e5d2ae491cd805be78038493d437dd1f7f3cc
* | | | | Merge "Add public network support when launching an instance."Jenkins2012-09-032-35/+117
|\ \ \ \ \
| * | | | | Add public network support when launching an instance.Akihiro MOTOKI2012-08-232-35/+117
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1039419. A concept of public network has been implemented in Quantum. To launch an instance connected to public network, we need to check public network (whose 'shared' attribute is True) in addition to network owned by the current tenant. Change-Id: I128e68a8b9404056f74153bf8f576cfa8b438e19
* | | | | | Merge "Add extensions for flavor swap and rxtx_factor"Jenkins2012-09-0321-57/+449
|\ \ \ \ \ \
| * | | | | | Add extensions for flavor swap and rxtx_factorVishvananda Ishaya2012-08-3121-57/+449
| | |_|_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The swap and flavor attributes of a flavor are not in the spec. This moves the properties so they are generated by extensions. The output will not be changed if all extensions are enabled (by default), but we now have a way to document these extra attributes and disable them. DocImpact Change-Id: Iee1cb1b1ee4116a38b90db581c38468d3d92afaf
* | | | | | Merge "Add api sample tests for flavors endpoint"Jenkins2012-09-0317-0/+523
|\| | | | |
| * | | | | Add api sample tests for flavors endpointVishvananda Ishaya2012-08-3117-0/+523
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Adds samples for GET /flavors and GET /flavors/1 Change-Id: I864996a5f92fa8a553c013e0df09bfdf8f04c659
* | | | | | Merge "Fix quota reservation expiration"Jenkins2012-09-033-0/+27
|\ \ \ \ \ \
| * | | | | | Fix quota reservation expirationMark McLoughlin2012-09-013-0/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #1040942 The db.reservation_expire() function assumes a 'usage' attribute on Reservation objects, but we don't actually define that relationship. The end result is that reservation_expire() currently traceback if it actually needs to expire any reservations. This happens pretty rarely since reservations should only need expiring if they are leaked because of another bug. Also define a test case to actually excercise the expiration code path. Add a missing chain-up to tearDown in test_limits which was causing the get_project_quotas() stub not to be unset and, in turn, the reservation expiration test to fail. Change-Id: Ib61dbf9fd5dfb5badaf05f20c423a69925d83754
* | | | | | | Merge "Clear up the .gitignore file"Jenkins2012-09-011-14/+14
|\ \ \ \ \ \ \ | |_|/ / / / / |/| | | | | |
| * | | | | | Clear up the .gitignore filepyw2012-08-311-14/+14
| | |_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 1.Remove duplicate items 2.Sort to alphabetical order 3.Remove "*.sw[po]" instead "*.swp" and "*.swo" for pass bundling test. Fixes: bug #1039898 Change-Id: Id8e03ecce6d2ab6d823a992a3a225b5ceb44aa80
* | | | | | Merge "Address race condition from concurrent task state update"Jenkins2012-08-316-38/+187
|\ \ \ \ \ \
| * | | | | | Address race condition from concurrent task state updateYun Mao2012-08-316-38/+187
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | task_state acts like a guard to avoid concurrent tasks to be scheduled. There might be two race conditions: 1) two tasks are concurrently accepted by api, check task_state to be None and allow the tasks to be executed concurrently. 2) one ordinary task is running, so that task_state is set. The delete task is accepted at API, and will "take over" and change task_state to DELETING. However the first task may continue to update task_state or set it to None as it finishes. This patch specifies current expected task_state when updating task_state. If unexpected state is met, abort the task without updating. Various compute tests are fixed accordingly to set the pre condition of the task state. Part of bug 1037372 Change-Id: I5fdf0946c728a47febb56ad468043a828b2736c8
* | | | | | | Merge "Avoid leaking security group quota reservations"Jenkins2012-08-312-5/+38
|\ \ \ \ \ \ \
| * | | | | | | Avoid leaking security group quota reservationsMark McLoughlin2012-08-312-5/+38
| | |_|/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #1043765 If you attempt to create a security group where one already exists using that name, a security groups quota reservation is leaked. Fix by ensuring the reservation is rolled back in all failure cases. Add a check to the tests to ensure that a reservation isn't leaked in any of our sad path tests for security group creation. Change-Id: I22d6e4a0452d3ac7a23056b0a4aee6f7dd135e30
* | | | | | | Merge "Fix deallocate_fixed_ip invocation"Jenkins2012-08-313-1/+22
|\ \ \ \ \ \ \
generated by cgit (git 2.34.1) at 2026-03-30 15:03:53 +0000