summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
| * | | | | | | | Introduce ImagePropertiesFilter scheduler filterJim Fehlig2012-08-177-224/+146
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit d39137fa added functionality to the ComputeFilter to filter on architecture, hypervisor_type, and vm_mode. The existing ArchFilter already filtered on architecture. This patch merges the ComputeFilter functionality introduced in d39137fa with the ArchFilter to create the ImagePropertiesFilter. The ImagePropertiesFilter uses image properties specified in the request_spec to filter hosts. This patch also adds the ImagePropertiesFilter to the list of default filters specified by the scheduler_default_filters FLAG. Fixes LP Bug #1037339 DocImpact Change-Id: Ifa6fccf2db266b0fe3457d58fc81fdb50ffcbc0f
* | | | | | | | | Merge "Debugged extra_specs_ops.py"Jenkins2012-08-212-2/+28
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Debugged extra_specs_ops.pyJinwoo 'Joseph' Suh2012-08-212-2/+28
| | |/ / / / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The current extra_specs_ops.py under nova/scheduler/filters has a bug: it crahses if <or> is not followed by a word. For example, if "<or> gpu <or>" is given in extra_specs, current code tries to pop a word after the second <or> that causes a crash. This patch lets it ignore the "<or>" keyword if no word follows the second "<or>" keyword. Debugged bug 1038979 Change-Id: I80a6eafd7ee57a2f3ccbf2fb869655ab84e92cf2
* | | | | | | | | Merge "Add deprecated warning to SimpleScheduler"Jenkins2012-08-211-0/+12
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Add deprecated warning to SimpleSchedulerJoe Gordon2012-08-211-0/+12
| | |_|_|_|_|_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1037829 Change-Id: I83abd3b58bc7d253616ad355458214859bb4112d
* | | | | | | | | Merge "Have compute_node_get() join 'service'"Jenkins2012-08-211-9/+10
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Have compute_node_get() join 'service'Chris Behrens2012-08-211-9/+10
| |/ / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1039633 compute_node_get() is used by hypervisors extension and it accesses compute_node['service']. Lazy joins fail now and then.. Also fixes line continuation indents with other compute_node calls in sqlalchemy/api.py Change-Id: I3ac02aad739e96a4866a64cabe859cf484341784
* | | | | | | | | Merge "Cleanup pip dependencies"Jenkins2012-08-212-2/+1
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Cleanup pip dependenciesJoe Gordon2012-08-212-2/+1
| | |_|_|_|_|_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix bug 1039745 * move feedparser to tools/test-requires * remove python-daemon Change-Id: I8da6ee6854e2064b44f3161b84f28d0bfc7f438d
* / | | | | | | | Remove unused exceptions from nova/exception.pyJoe Gordon2012-08-211-130/+0
|/ / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Removed exceptions: ImagePaginationFailed MelangeConnectionFailed InvalidSignature InvalidInstanceType InstanceNotSuspended DestinationHostUnavailable SourceHostUnavailable VolumeNotFoundForInstance NoVolumeTypesFound ListingImageRefsNotSupported KernelNotFoundForImage ProjectMembershipNotFound UserRoleNotFound NetworkBusy NetworkIsDuplicated FixedIpNotFoundForHost AuthTokenNotFound AccessKeyNotFound NoInstanceTypesFound LDAPUserNotFound LDAPGroupNotFound LDAPGroupMembershipNotFound GlobalRoleNotAllowed LDAPUserExists LDAPGroupExists LDAPMembershipExists ProjectExists ZoneRequestError SolidFireAPIStatusException LDAPObjectNotFound UserExists UserNotFound Change-Id: I8a3caaebdeed82f19733893e62f57648bdfc45bd
* | | | | | | | Merge "Make glance image service check base exception classes"Jenkins2012-08-212-14/+53
|\ \ \ \ \ \ \ \
| * | | | | | | | Make glance image service check base exception classesChris Behrens2012-08-212-14/+53
| |/ / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1039675 glanceclient can raise HTTPNotFound (as well as NotFound) it appears, but glance image service is only converting NotFound -> InstanceNotFound. Same applies to 'Forbidden' and other exceptions. This patch converts 'exc_type is NotFound'-like checks to use 'isinstance' instead, which will cover HTTPNotFound, etc. Change-Id: I0982875e667121cee3da3cfe0124499cad484fa6
* | | | | | | | Merge "Implements sending notification on metadata change"Jenkins2012-08-214-2/+30
|\ \ \ \ \ \ \ \ | |/ / / / / / / |/| | | | | | |
| * | | | | | | Implements sending notification on metadata changeNikhil Komawar2012-08-214-2/+30
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It would be good to have a notification sent every time there is change in the metadata. This merge prop implements sending a compute.instance.update event for the above change. fixes bug 1034547 Change-Id: I451796b6a83c4782df6f72f083755527ebd62a52
* | | | | | | Merge "Remove stub_network flag"Jenkins2012-08-2112-43/+66
|\ \ \ \ \ \ \
| * | | | | | | Remove stub_network flagRongze Zhu2012-08-2112-43/+66
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #816025 remove stub_network flag and replace usage with stubbing out of relevant calls in tests. In nova/tests/fake_network.py , I add set_stub_network_methods and unset_stub_network_methods. They can be userd for tests. Change-Id: Ie8fa95d9ab62cced00f0917cffe75ebc6c94bb3e
* | | | | | | Merge "Keep the ComputeNode model updated with usage"Jenkins2012-08-2129-301/+1675
|\ \ \ \ \ \ \ | |/ / / / / / |/| | | | | |
| * | | | | | Keep the ComputeNode model updated with usageBrian Elliott2012-08-2029-301/+1675
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Keep the compute host's ComputeNode model in sync with the level of resource usage. This enables the ComputeNode model to be used as a basis for scheduling decisions rather than forcing scheduler to calculate free resources from an instance_get_all on each request. Resources like memory and disk are claimed as instances are built or deleted. There is also support for configurable compute node stats (a generic key/value store) for extensible advertising of other usage stats that may be useful for a particular scheduler implementation. Additionally, there is a periodic task on the compute host that audits actual resource consumption at the virt layer to ensure that the database stays in sync. This change partially implements blueprint: scheduler-resource-race This patch complements: https://review.openstack.org/#/c/9540/ (build re-scheduling support) Change-Id: Ibbe3839a054f8b80664b413d47f766ca8d68e3f2
* | | | | | | Merge "Return HTTP 422 on bad server update PUT request."Jenkins2012-08-213-2/+23
|\ \ \ \ \ \ \
| * | | | | | | Return HTTP 422 on bad server update PUT request.Dan Prince2012-08-173-2/+23
| | |/ / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch updates the /servers API and the disk config extension so that they properly handle incorrectly formatted server update/PUT requests. Includes a new test cause that verifies (with extensions enabled) that a poorly formatted PUT request to /servers returns HTTP 422 instead of a 500. Also updated the previous test_create_missing_server test to use test.TestException instead of Exception. Fixes LP Bug #1038227. Change-Id: I6b11602790778465348647bb5825b7326d50065a
* | | | | | | Merge "Makes sure instance deletion ok with deleted data"Jenkins2012-08-213-21/+58
|\ \ \ \ \ \ \
| * | | | | | | Makes sure instance deletion ok with deleted dataVishvananda Ishaya2012-08-173-21/+58
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 5ad1dea4 added changed the network deallocation code to work with deleted instances. This was done by setting the context to read deleted records. Unfortunately this was done a little too broadly, leading to a new bug where a deleted floating_ip will cause an instance to not be able to be deleted. This fixes the issue by limiting the use of read_deleted context to only the places it is trying to access the instance record. It adds a test to verify that the code works with a duplicate deleted floating_ip and updates the existing test for a deleted instance to exercise the entire code path. Fixes bug 1038266 Change-Id: I1aef94369e5bcf951e78e89b1eded5305cf36b53
* | | | | | | Merge "OpenStack capitalization added to HACKING.rst"Jenkins2012-08-2121-22/+31
|\ \ \ \ \ \ \
| * | | | | | | OpenStack capitalization added to HACKING.rstJoe Gordon2012-08-1721-22/+31
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Along with capitalization fixes to comments in code Change-Id: I72ddc582001f80d954ca5a121903c689f40d08d1
* | | | | | | Merge "Allow network to call get_fixed_ip_by_address."Jenkins2012-08-212-0/+2
|\ \ \ \ \ \ \ | |_|_|/ / / / |/| | | | | |
| * | | | | | Allow network to call get_fixed_ip_by_address.Dan Prince2012-08-202-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Updates the default Nova policy.json so that the network API can call get_fixed_ip_by_address. This was something we identified when c6be78deb6a3a7cbf092efdfb54f561ce689605f caused some regressions which prohibited the Nova metadata server from making this call. We have subsequently fixed the metadata server by correcting admin context access to this function.... however using an explicit policy rule (and not relying on the default) seems like a good idea as well. Partial fix for LP Bug #1039093. Change-Id: I88d0f5a4dc263bf66f316a4d26c734b6bc9dad21
* | | | | | | Merge "Add key_name attribute in XML servers API"Jenkins2012-08-201-1/+1
|\ \ \ \ \ \ \
| * | | | | | | Add key_name attribute in XML servers APIMauro S. M. Rodrigues2012-08-201-1/+1
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1039139 The API was ignoring the 'key_name' attribute and so it was impossible to create a server associated with a keypair. This patch fix the issue. Change-Id: I1efe542081382f3a19b31c09ddfc93929b1485d0
* | | | | | | Merge "Fix PEP8 issues"Jenkins2012-08-202-27/+36
|\ \ \ \ \ \ \ | |/ / / / / / |/| | | | | |
| * | | | | | Fix PEP8 issuesRay Chen2012-08-172-27/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix some PEP8 issues in doc/ext/nova_todo.py and doc/source/conf.py and make the code look more clearly. Change-Id: I2b0ce1b09a4a707cffaa565747aabd5346eb9f41
* | | | | | | Fix is_admin check via policyVishvananda Ishaya2012-08-204-7/+12
| |_|/ / / / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we create an admin context internally, we need to make sure that all of the policy actions that are available to that context are still available. This wasn't working in some cases because we were looking for a hard-coded role called 'admin'. Fixes bug 1039093 Change-Id: I939f834a63c9e6e6e7c87b115bd469466da66a69
* | | | | | Merge "Move volume size validation to api layer."Jenkins2012-08-204-18/+67
|\ \ \ \ \ \
| * | | | | | Move volume size validation to api layer.Eoghan Glynn2012-08-154-18/+67
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As per the discussion on this patch: https://review.openstack.org/10540 Change-Id: I5d3031a08aa1e2d17b5fb233ebaec938c66fa88b
* | | | | | | Merge "Fix spelling typos"Jenkins2012-08-2011-18/+18
|\ \ \ \ \ \ \
| * | | | | | | Fix spelling typosJoe Gordon2012-08-1611-18/+18
| | |_|_|_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes for: * nova/*py * nova/scheduler/ * nova/compute/*py * nova/api/ec2/ Change-Id: I7bec89b3551eda84d596e55eeba9ccb9d944a5cb
* | | | | | | Merge "Remove unused function in console api"Jenkins2012-08-201-6/+0
|\ \ \ \ \ \ \
| * | | | | | | Remove unused function in console apiSulochan Acharya2012-08-151-6/+0
| | |_|_|_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove _translate_id_if_necessary from nova.console.api since its now using instance_uuid and this function is not used anymore. Complete fix for bug 1014796. Change-Id: I1350497541ca6571b38a05d66ca3c5498d63a9b5
* | | | | | | Merge "Refactor extra specs matching into a new module"Jenkins2012-08-202-45/+65
|\ \ \ \ \ \ \
| * | | | | | | Refactor extra specs matching into a new moduleMark McLoughlin2012-08-172-45/+65
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We want AggregateInstanceExtraSpecsFilter to have the same extra specs matching semantics as ComputeCapabilitiesFilter, so split the code out so that it can be reused. Change-Id: I6938c70f7272956eb27a6609546c3e20ec0dd99e
* | | | | | | | Merge "Disable I18N in Nova's test suites"Jenkins2012-08-202-0/+8
|\ \ \ \ \ \ \ \
| * | | | | | | | Disable I18N in Nova's test suitesJian Wen2012-08-182-0/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Messages translated into other language like Chinese cause some unit tests to fail. Fixes LP bug #1037946. Change-Id: I01d04333d28bab52d6f424432524b206f3845fba
* | | | | | | | | Merge "Make missing imports flag in hacking settable."Jenkins2012-08-201-4/+5
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Make missing imports flag in hacking settable.Monty Taylor2012-08-171-4/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change-Id: Ib1ada320588eb42370c9fa2c0eff4eb888258ea5
* | | | | | | | | | Merge "Restore libvirt block storage connections on reboot."Jenkins2012-08-208-19/+64
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Restore libvirt block storage connections on reboot.Rafi Khardalian2012-08-148-19/+64
| | |_|_|_|_|_|_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1036902. There are a number of cases where block storage connections are not properly restored, impacting libvirt in particular. The most common case is a VM which has block storage attached via iSCSI, whereby the physical system is rebooted. When the system comes back up and starts nova-compute, the iSCSI connections are NOT recreated for the instances slated to be resumed (assuming resume_guests_state_on_host_boot is set). The patch changes the compute manager to pass block_storage_info via driver.reboot() and driver.resume_state_on_host_boot(). The fix is actually only present in the libvirt driver. However, all the other drivers were updated to accept the additional, optional function arg. With the changes in place, iSCSI connections for libvirt are re-established after a hypervisor reboot with resume_guests_state_on_host_boot=True and on every hard_reboot. The latter is intended so that users have a last ditch option for recovering their VMs without administrative involvement. Change-Id: Idf5d53f21991a359bec6ce26ae9fe3bd61800ce3
* / | | | | | | | | Remove hard-coded 'admin' role checking and use policy insteadunicell2012-08-196-3/+35
|/ / / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bug 1037786 This change removed hard-coded 'admin' role checking, and use policy check to decide is_admin at the time of context construction. And also set default admin role to 'admin', in policy template, to keep backward compatibility. Change-Id: I56b45bfdfba1b8e1391bde7fec3d4cf26c66324c
* | | | | | | | | Merge "Fix get_vnc_console race."Jenkins2012-08-172-0/+12
|\ \ \ \ \ \ \ \ \
| * | | | | | | | | Fix get_vnc_console race.Russell Bryant2012-08-172-0/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There exists a window between a server create is done and when the host is assigned by the scheduler that a request to get the vnc console will result in a 500 error. This patch adds a check in the compute API to ensure that the instance has a host assigned before try to execute a rpc on the compute node. If not, it raises InstanceNotReady (which translates to a 409 in the OS API). Fix bug 1037809. Change-Id: I3f5af90d57ae84f98f787e14ccb66f1841ac0c6d
* | | | | | | | | | Merge "Revert to prior method of executing a libvirt hard_reboot."Jenkins2012-08-171-6/+2
|\ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | Revert to prior method of executing a libvirt hard_reboot.Rafi Khardalian2012-08-141-6/+2
| |/ / / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1036826. Using reset for hard_reboot is not reliable, even where it is supported by libvirt. Hard reboots are one of the only ways to recover a VM in a broken state. The reset command assumes the domain is running in some capacity and will fail if it is not. Hard reset is the current sledgehammer for fixing issues and it really needs to stay that way. Change-Id: I84705b72d79cf71adad066b18267fdfb199bc9cb
generated by cgit (git 2.34.1) at 2026-03-31 06:01:00 +0000