summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
| * | | | | | | | | | Fix typo in inherit_properties_from_imageChuck Short2012-11-291-1/+1
| | |_|_|_|_|_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Should be architecture rather than arch. Fixes LP: 1082547 Change-Id: I9e40b00ec05a99f11894b0e47a9cf47b20b46fe2 Signed-off-by: Chuck Short <chuck.short@canonical.com>
* | | | | | | | | | Merge "Complete API samples for Hosts extension"Jenkins2012-11-2921-0/+96
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Complete API samples for Hosts extensionMauro S. M. Rodrigues2012-11-2221-0/+96
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The API samples for Hosts extension missed power operations (shutdown, reboot and startup) and maintenance/status operations. Complete fix to bug 1070167 Change-Id: I07fd34e071eae9afc82f9320634df24f48623e0b
* | | | | | | | | | | Merge "Fix HostDeserializer to enable multiple line xml"Jenkins2012-11-292-8/+19
|\| | | | | | | | | |
| * | | | | | | | | | Fix HostDeserializer to enable multiple line xmlMauro S. M. Rodrigues2012-11-222-8/+19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug: 1082132 The current HostDeserializer wasn't able to deal with multiple line xml inputs. Since the HostDeserializer is used only on the update operation this patch renames it to HostUpdateDeserializer, and parses only the valid options (status and maintenance_mode). Change-Id: I9ec3b296a07e0fc903da1c005cf7bd5dcb08fd1b
* | | | | | | | | | | Merge "Fix args to AggregateError exception."Jenkins2012-11-291-1/+2
|\ \ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | | Fix args to AggregateError exception.Russell Bryant2012-11-291-1/+2
| | |/ / / / / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds some arguments to the AggregateError exception when raised from a test. This didn't actually break the test, but it threw me off for a little bit. I had done something to break this test, so when I went to look at why, there was an exception logged because of this mistake, which wasn't related to what the real problem was. To avoid this unnecessarily making it into the test log, pass some stuff so that building the exception message succeeds. Change-Id: I4460f4844c16877ef78b363842518b0ba96eb9aa
* | | | | | | | | | | Merge "Remove start_guests_on_host_boot config option"Jenkins2012-11-292-8/+1
|\ \ \ \ \ \ \ \ \ \ \ | |_|_|_|_|_|/ / / / / |/| | | | | | | | | |
| * | | | | | | | | | Remove start_guests_on_host_boot config optionRafi Khardalian2012-11-292-8/+1
| | |_|/ / / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1081808 The start_guests_on_host_boot config option ignores any/all user-requested VM state and attempts to start every VM assigned to the hypervisor in question. This means even a VM which was suspended, paused, in error state, etc will run through the process of attempting to start. resume_guests_state_on_host_boot provides similar functionality, except that it confirms the VM was previously in a running state prior to attempting to restart on compute startup. This patch will remove the config option entirely. Flags: DocImpact Change-Id: Ib26e74c46e17496bccf8e6fa3002fc4c78bde9e9
* | | | | | | | | | Merge "Fix KeyError of log message in compute/api.py"Jenkins2012-11-291-1/+1
|\ \ \ \ \ \ \ \ \ \ | |_|_|_|_|_|/ / / / |/| | | | | | | | |
| * | | | | | | | | Fix KeyError of log message in compute/api.pyWangpan2012-11-291-1/+1
| | |/ / / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove the NOT exist parameter 'msg' in log. Fixes: bug #1084335 Change-Id: I8bc7a32edcff5e61505f2bc564bae916a753a97f
* | | | | | | | | Merge "Don't run the periodic task if ticks_between_runs is below zero"Jenkins2012-11-291-2/+3
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Don't run the periodic task if ticks_between_runs is below zeroDavanum Srinivas2012-11-291-2/+3
| |/ / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | if the user sets say image_cache_manager_interval=-1, then don't add the task to the list of periodic tasks remove extra braces that i had added DocImpact Fixes LP #1084232 Change-Id: Ieecd67ddbc70b815a88f40e72ca2899787d75988
* | | | | | | | | Merge "added getter methods for quantumv2 api"Jenkins2012-11-291-2/+4
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | added getter methods for quantumv2 apiChristian Berendt2012-11-131-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change-Id: Ie2b7ba698c232f55e112b978d7a548b48462fca3
* | | | | | | | | | Merge "Access DB values as dict not as attributes"Jenkins2012-11-295-41/+42
|\ \ \ \ \ \ \ \ \ \ | |_|/ / / / / / / / |/| | | | | | | | |
| * | | | | | | | | Access DB values as dict not as attributesJoe Gordon2012-11-285-41/+42
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We cannot assume nova.db.api will be returning sqlalchemy objects with attributes, instead treat return values as dicts Part of blueprint db-api-cleanup Change-Id: Icb955118aabff1d8f9ee7c3e95f8f2b3ffc671ee
* | | | | | | | | | Merge "Fix import problem in test_virt_disk_vfs_localfs."Jenkins2012-11-291-3/+3
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Fix import problem in test_virt_disk_vfs_localfs.Russell Bryant2012-11-281-3/+3
| | |_|/ / / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | jog0 pointed out that you couldn't run this test, at least directly, using tox. This patch fixes the error. Change-Id: I72e81a9931a1601085932273b61823c32e1472aa
* | | | | | | | | | Merge "Call plug_vifs() for all instances in init_host"Jenkins2012-11-291-5/+4
|\ \ \ \ \ \ \ \ \ \ | |/ / / / / / / / / |/| | | | | | | | |
| * | | | | | | | | Call plug_vifs() for all instances in init_hostMorgan Fainberg2012-11-281-5/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove the assumption that we are using the "flat" networking model and there is a single bridge interface for all instances on a given compute node. Instead simply call plug_vifs() for each instance in init_host. Fixes bug 1083784 Change-Id: I4b367deb6851c06c80a4bdf0bd806cf5e867bb0a
* | | | | | | | | | Merge "adjust rootwrap filters for recent file injection changes"Jenkins2012-11-291-30/+15
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | adjust rootwrap filters for recent file injection changesPádraig Brady2012-11-221-30/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Removed filters for guestmount, fusermount and cp. Adjusted comment paths as per the recent refactoring. Change-Id: I643554b89862225d16456e47daf6c11828f602d2
* | | | | | | | | | | Merge "Make compute manager use conductor for instance_gets"Jenkins2012-11-295-8/+55
|\ \ \ \ \ \ \ \ \ \ \ | |_|_|/ / / / / / / / |/| | | | | | | | | |
| * | | | | | | | | | Make compute manager use conductor for instance_getsDan Smith2012-11-285-8/+55
| | |_|_|_|_|/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds support to nova-conductor for the following operations: instance_get_by_uuid() instance_get_all_by_host() It also makes the compute/manager use nova-conductor for those operations. Change-Id: Ia872a2ec6061b4aac9f89366125060fc0844dc22
* | | | | | | | | | Merge "Convert datetimes for conductor instance_update."Jenkins2012-11-293-3/+17
|\ \ \ \ \ \ \ \ \ \ | |_|_|_|_|_|/ / / / |/| | | | | | | | |
| * | | | | | | | | Convert datetimes for conductor instance_update.Russell Bryant2012-11-283-3/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch ensures that jsonutils.to_primitive() is used on the updates being sent to conductor's instance_update method. This is primarily so that we convert datetimes to a string using a very specific format defined in the timeutils module. Then, in the conductor manager, convert these fields back into datetime objects using timeutils. The conversion to a string was already happening implicitly if you were using the kombu rpc driver. The reason is that kombu uses anyjson, which we hook in to and will automatically use our own to_primitive() to fix serialization problems for kombu. Doing it explicitly is better, and is required for the other rpc drivers to work. This is to address a problem with sqlite where it would raise an exception when we try to do an update to a datetime field with a string. Fix bug 1084240. Change-Id: I2d703a92df1e85f0f3a7a793b301fe16cbd81ff5
* | | | | | | | | | Merge "Fixes HyperV compute "resume" tests"Jenkins2012-11-281-8/+2
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Fixes HyperV compute "resume" testsAlessandro Pilotti2012-11-291-8/+2
| |/ / / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes Bug #1083018 Adds support for the extra parameter added in the compute driver's resume method. The added network_info parameter is currently not used and can safely be set to None. Change-Id: I887d48f39601323433384833f8b8d33132858b59
* | | | | | | | | | Merge "Include 'hosts' and 'metadetails' in aggregate."Jenkins2012-11-283-7/+31
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Include 'hosts' and 'metadetails' in aggregate.Russell Bryant2012-11-283-7/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When converting an Aggregate sqlalchemy model to primitive types using jsonutils.to_primitive(), the hosts and metadata were being left out. This is because they are implemented as properties on the model and must be explicitly listed as extra keys to be read. We also want to make sure that the _hosts and _metadata are eager loaded. This patch ensures this gets done by adding it to the _get_aggregate_query() helper, and making sure it is used everywhere that is appropriate. Also, on creation of the aggregate, just manually set _hosts and _metadata to empty lists, since we know they're empty at this point. These changes caused a side effect to the os-aggreages API extension. The result from creating an aggregate started including 'hosts' and 'metadata'. There's a hack in there to maintain the same API result. These don't add any value to this API call anyway, since we know they're blank at this point. Part of blueprint no-db-compute. Change-Id: I4048ea6f477f7a695701eb3ba57f97afca0c0e41
* | | | | | | | | | | Merge "Verify doc/api_samples files along with the templates"Jenkins2012-11-2811-59/+171
|\ \ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | | Verify doc/api_samples files along with the templatesDan Smith2012-11-2811-59/+171
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This also includes a bunch of updates to the actual samples that have become stale as things have changed in the tree and naughty developers haven't updated the docs. This new bit of logic in the samples test should help prevent that in the future. In order to verify the static doc/api_samples files, this adds a new method to the base test, which allows the subclasses to refine their subs into pattern matches after the server response is verified, and before the doc/api_samples file is verified. Change-Id: Ie46b24a65145c407d46974b63a97a84911223547
* | | | | | | | | | | | Merge "Allow multi_host compute nodes to share dhcp ip"Jenkins2012-11-286-4/+152
|\ \ \ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | | | Allow multi_host compute nodes to share dhcp ipVishvananda Ishaya2012-11-266-4/+152
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds a new flag: share_dhcp_address which if enabled in multihost mode will allow all compute nodes to share an ip on guest network. The code will isolate the address using iptables and ebtables so it is only visible to the vms. This patch has two benefits: a) we don't have to use an ip address from every network for each compute node. This is especially valuable in vlan mode where the networks are generally small b) we can improve security by blocking all access to the ip on the guest network from outside the compute node. While we could do similar blocking using a different ip for each node, it makes dhcp setup much more complicated if a vm is migrated to another node. Implements blueprint shared-dhcp-ip Change-Id: Iaf84c0ad2848921122866956105eb44c074450dc
* | | | | | | | | | | | | Merge "Update migration so it supports PostgreSQL."Jenkins2012-11-281-2/+2
|\ \ \ \ \ \ \ \ \ \ \ \ \ | |_|_|_|/ / / / / / / / / |/| | | | | | | | | | | |
| * | | | | | | | | | | | Update migration so it supports PostgreSQL.Dan Prince2012-11-281-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Updates the 'deleted' columns so that they are treated as Booleans instead of integers. This makes it so that 144_add_node_to_migrations.py works on both MySQL and PostgreSQL. Fixes LP Bug #1084208. Change-Id: Ibff1c45117e47126a255b2a1df5e915c881cb5c1
* | | | | | | | | | | | | Merge "Add methods to Host operations to fake hypervisor"Jenkins2012-11-281-3/+7
|\ \ \ \ \ \ \ \ \ \ \ \ \ | |/ / / / / / / / / / / / |/| | | | | | | | / / / / | | |_|_|_|_|_|_|/ / / / | |/| | | | | | | | | |
| * | | | | | | | | | | Add methods to Host operations to fake hypervisorMauro S. M. Rodrigues2012-11-211-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch implements some features present in os-hosts extension that are not implemented in fake hypervisor: - Power Actions: Shutdown/Reboot/Startup a host - Change host status: enable or disable - Set/Unset the host into maintenance mode With this patch we're able to get full api samples for this extension Change-Id: I3b76e3070e326b39a67b9d4d25b5e3d91a3c102f
* | | | | | | | | | | | Merge "Remove default_image config option"Jenkins2012-11-282-5/+1
|\ \ \ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | | | Remove default_image config optionMark McLoughlin2012-11-282-5/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The value of this option hasn't been used outside of the tests since the very early days of the project. There's no particular reason to have a config option for this test value, so just delete it. blueprint: scope-config-opts Change-Id: I00a08d4dc539f92f4fa4bd8c3c8756e5029d193f
* | | | | | | | | | | | | Merge "Use flags() helper method to override config in tests"Jenkins2012-11-281-2/+2
|\ \ \ \ \ \ \ \ \ \ \ \ \ | |_|_|_|_|_|_|_|_|_|_|/ / |/| | | | | | | | | | | |
| * | | | | | | | | | | | Use flags() helper method to override config in testsMark McLoughlin2012-11-281-2/+2
| | |_|_|_|_|/ / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Setting config option values using setattr() is dangerous because the value remains set even after the particular test finishes, so you can end up with other tests relying on it and being unable to run tests in isolation. We have a helper method for doing this, just use it. Change-Id: I872d7a4f70eceb9ff64f56901cb45552f237c733
* | | | | | | | | | | | Merge "Move ec2 config opts to nova.api.ec2.cloud"Jenkins2012-11-283-26/+25
|\ \ \ \ \ \ \ \ \ \ \ \ | | |/ / / / / / / / / / | |/| | | | | | | | | |
| * | | | | | | | | | | Move ec2 config opts to nova.api.ec2.cloudMark McLoughlin2012-11-283-26/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Only ec2_dmz_host and ec2_port are used outside of ec2.cloud and in both cases pipelib is the other module that uses them. Move all the ec2 options to ec2.cloud and make pipelib explicitly depend on the ones it needs. blueprint: scope-config-opts Change-Id: Ib187ec3421a25bdd86b13dfeae1dc9a07873eb43
* | | | | | | | | | | | Merge "Move imagecache code from nova.virt.libvirt.utils"Jenkins2012-11-285-163/+150
|\| | | | | | | | | | |
| * | | | | | | | | | | Move imagecache code from nova.virt.libvirt.utilsMark McLoughlin2012-11-285-163/+150
| | |_|_|_|/ / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The imagecache related code in libvirt.utils uses the base_dir_name config option. If we move the imagecache utils into imagecache, then we can move the base_dir_name option into that module since there are no other users of it. Having all the imagecache code in the imagecache module seems like a sensible thing to do anyway. blueprint: scope-config-opts Change-Id: I1e154aa4de1628d40964207bdcbe5b3b55076442
* | | | | | | | | | | Merge "Make resize and multi-node work properly together"Jenkins2012-11-2811-40/+71
|\ \ \ \ \ \ \ \ \ \ \ | |/ / / / / / / / / / |/| | | | | | | | | |
| * | | | | | | | | | Make resize and multi-node work properly togetherBrian Elliott2012-11-2711-40/+71
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Added node support to migrations and update node properly in 'resize_instance'. bug 1081355 Change-Id: I003d34e3f7ed9ce2feda19ee5ce210ed4ba7eaa1
* | | | | | | | | | | Merge "Migration model update for multi-node resize fix."Jenkins2012-11-282-0/+188
|\| | | | | | | | | | | |/ / / / / / / / / |/| | | | | | | | |
generated by cgit (git 2.34.1) at 2026-02-07 17:06:10 +0000