summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* | | | | | | | | Fix a bug in compute_node_statisticsKevin L. Mitchell2012-07-241-1/+2
|/ / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Turns out, using Sqlalchemy's sum aggregation functions returns instances of Decimal. This caused the statistics reporting to fail with a cryptic "Circular reference" error during serialization. Fixed by forcing the values to integer. Change-Id: Ia89c191f420e116b2f7ca144cdb2e21e6ace404a
* | | | | | | | Merge "Updates migration 111 to work w/ Postgres."Jenkins2012-07-241-0/+5
|\ \ \ \ \ \ \ \
| * | | | | | | | Updates migration 111 to work w/ Postgres.Dan Prince2012-07-211-0/+5
| |/ / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Updates migration 111 so that we drop the aggregate_hosts_host_key unique constrain by name. Fixes LP Bug #1027453 Change-Id: Ibe36eace6a2d5d5a8d8d03d2cb170cbe0056320e
* | | | | | | | Merge "Add git commit message validation to hacking.py"Jenkins2012-07-241-1/+29
|\ \ \ \ \ \ \ \ | |_|_|_|_|_|/ / |/| | | | | | |
| * | | | | | | Add git commit message validation to hacking.pyJoe Gordon2012-07-201-1/+29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Adds N801: git commit title should provide a description Change-Id: Ic6334b20f5f06631b7a5579a4d1b2303685463eb
* | | | | | | | Xen: Fix snapshots when use_cow=True.Rick Harris2012-07-241-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes an issue where use_cow=True would sporadically cause snapshot corruption. The root cause was a race-condition where the VM's sm_config was sometimes out-of-sync with the underlying VHDs. The solution is to force an SR rescan immediately after taking a snapshot so that any VHD parent-pointer modifications are propogated into the sm_config so that _walk_vdi_chain will receive accurate information. Change-Id: I03b4914d9420918590ae27dfc77dfedac9c767c1
* | | | | | | | Merge "Add call to get hypervisor statistics"Jenkins2012-07-244-3/+130
|\ \ \ \ \ \ \ \ | |_|_|_|_|_|/ / |/| | | | | | |
| * | | | | | | Add call to get hypervisor statisticsKevin L. Mitchell2012-07-234-3/+130
| | |_|/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Adds an admin API call to retrieve the compute node statistics for an entire nova instance. Counts up all hypervisors and sums all their values (vcpus, vcpus_used, etc.). Change-Id: I36272656bb417c1549133fd2963bac54db0e86c6
* | | | | | | Merge "Use compute_driver in vmware driver help messages."Jenkins2012-07-231-6/+10
|\ \ \ \ \ \ \
| * | | | | | | Use compute_driver in vmware driver help messages.Dan Prince2012-07-231-6/+10
| |/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Updates the help messages in the vmware driver to use 'compute_driver' examples instead of 'connection_type' which is now deprecated. Change-Id: Iddea2cabc841200edd561d71f0c11fd6ba272f5d
* / / / / / / Update NovaKeystoneContext to use jsonutils.Dan Prince2012-07-231-2/+2
|/ / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Minor fix to use jsonutils.loads instead of json.loads to standardize w/ the rest of Nova. Change-Id: I4a6912d29fbe425b29bd1b170b9bec19e3bfc46e
* | | | | | Merge "Log snapshot UUID and not OpaqueRef."Jenkins2012-07-231-6/+4
|\ \ \ \ \ \
| * | | | | | Log snapshot UUID and not OpaqueRef.Rick Harris2012-07-231-6/+4
| | |_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The OpaqueRef doesn't help troubleshooting since it can't be easily used to identify an image after the session finishes. UUID, which is invariant, is much more useful. Change-Id: I3b802e41e583bee84ee6f5d05e1ffc581761571d
* / | | | | Remove unused user_id and project_id argumentsJohannes Erdfelt2012-07-232-6/+5
|/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | The code using the arguments was removed a long time ago. Change-Id: Ic64ce46115617d05e41b6d01c8d403b58f8cacf9
* | | | | Merge "Update jsonutils from openstack-common."Jenkins2012-07-231-2/+4
|\ \ \ \ \
| * | | | | Update jsonutils from openstack-common.Russell Bryant2012-07-231-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit 9e1bd9d9313a9f324c5b7b02232e8bd2fd12ea8a Author: Russell Bryant <rbryant@redhat.com> Date: Wed Jul 18 16:47:34 2012 -0400 Add missing convert_instances arg. When calling jsonutils.to_primitive() recursively, the convert_instances argument should be passed along. This change fixes one place where it was not. commit 2d6f84742a3e8ea51ebbfb82cbeacefe97e199d5 Author: Russell Bryant <rbryant@redhat.com> Date: Wed Jul 18 16:15:52 2012 -0400 Track to_primitive() depth after iteritems(). Change jsonutils.to_primitive() to increase the recursion depth counter when calling to_primitive() on the result of iteritems() from the current element. Previously, the only time the counter was increased was when converting the __dict__ from an object. The iteritems() case risks cycles, as well. I hit a problem with this when trying to call to_primitive on an instance of nova.db.sqlalchemy.models.Instance. An Instance includes a reference to InstanceInfoCache, which has a reference back to the Instance. Without this change, to_primitive() would raise an exception for an Instance due to excessive recursion. Related to nova blueprint no-db-messaging. Change-Id: Iaa49ea08b406a38840e8a0b5466d48f2d3f7e840
* | | | | | Merge "Adding networking rules to vm's on compute service startup"Jenkins2012-07-231-1/+8
|\ \ \ \ \ \
| * | | | | | Adding networking rules to vm's on compute service startupDerek Higgins2012-07-231-1/+8
| | |_|/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug 1027105 Restarting nova-compute resulted in a new empty iptables chain "nova-compute-FORWARD" being create. This commit adds a single call to plug_vifs which ensures the appropriate filters (and bridge) are present on service startup. Change-Id: I51cbb56acbb244ddddd951a26d3168cf5a9c0314
* | | | | | Merge "use boto's HTTPResponse class for versions of boto >=2.5.2"Jenkins2012-07-231-3/+7
|\ \ \ \ \ \ | |_|_|_|/ / |/| | | | |
| * | | | | use boto's HTTPResponse class for versions of boto >=2.5.2Ionuț Arțăriși2012-07-231-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug: 1027984 Change-Id: I2101ddaa78ede90f39b55d28aab73e7f5a816af5
* | | | | | Merge "Make it clear subnet_bits is unused in ipam case"Jenkins2012-07-231-2/+1
|\ \ \ \ \ \
| * | | | | | Make it clear subnet_bits is unused in ipam caseMark McLoughlin2012-07-231-2/+1
| | |/ / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In NetworkManager.create_networks(), the subnet_bits variable is unused where 'ipam' is passed as a parameter - i.e. when the method is called from QuantumNovaIPAMLib. This is because in the ipam case we never carve up the IP range into multiple networks. Move the variable around so that its usage scope is more obvious. Change-Id: I2a1a2cb63f9864c15b33b30f1bf995a50a6ec067
* | | | | | Merge "Remove unused add_network_to_project() method"Jenkins2012-07-236-33/+5
|\ \ \ \ \ \
| * | | | | | Remove unused add_network_to_project() methodMark McLoughlin2012-07-236-33/+5
| |/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This method was added here: https://code.launchpad.net/~tr3buchet/nova/multi_nic/+merge/64767 but appears never to have been used. We can safely remove it and the 'force' parameter to network_associate() which only it uses. This removal from the network service RPC API would not warrant a major version number bump since nothing ever used this part of the API. Change-Id: Iba3cf7b61b629ba513a58a38886c9322818accb5
* | | | | | Merge "Return 404 when attempting to remove a non-existent floating ip"Jenkins2012-07-232-3/+17
|\ \ \ \ \ \ | |_|_|/ / / |/| | | | |
| * | | | | Return 404 when attempting to remove a non-existent floating ipDan Smith2012-07-232-3/+17
| | |_|_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes bug 993774 and returns 404 instead of 202 when trying to delete a floating ip from an instance to which it is not allocated. Change-Id: I9538a8a71cbbb9d19e9ebb76f93b180dbf619764
* | | | | Merge "Adds non_inheritable_image_properties flag."Jenkins2012-07-232-0/+8
|\ \ \ \ \ | |/ / / / |/| | | |
| * | | | Adds non_inheritable_image_properties flag.Rick Harris2012-07-202-0/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some image-properties should not be inherited from the instance when taking a snapshot because we want that functionality to only available for the base-image, not subsequent snapshots. As an example, we would like 'cache_in_nova' fast-cloning to be available just for base-images and not customer snapshots created off a base-image. This patch adds a new configuration which allows operators to, at runtime, select which image-properties should not be inherited by newly created snapshots. Change-Id: I8d7781b05ffd71a59ba69fb7c3df4616cba94cce
* | | | | Merge "Avoid unrecognized content-type message"Jenkins2012-07-231-2/+7
|\ \ \ \ \ | |_|_|/ / |/| | | |
| * | | | Avoid unrecognized content-type messageMark McLoughlin2012-07-231-2/+7
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes bug #904473 If the client does not include a Content-type header in its request (which is normal for e.g. GET and DELETE) then eventlet will pass us 'text/plain'. Because 'text/plain' isn't one of our supported request content types we log: Unrecognized Content-Type provided in request This can be an annoying red-herring for people debugging issues. The reason that eventlet defaults to text/plain is because it uses the mimetools.Message.gettype() method: Message.gettype() Return the message type [..] as specified in the Content-Type header. If no such header exists, return 'text/plain'. The wsgiref web server has the same behavior, whereas mod_wsgi will pass no content-type and twisted will pass an empty string. Make our code more robust by treating the empty string or text/plain as if no content-type header was supplied. Change-Id: Ide117c807db0dc0f5cbe974788b604b5e236800a
* | | | Merge "Migrations for deleted data for previously deleted instances."Jenkins2012-07-231-0/+67
|\ \ \ \
| * | | | Migrations for deleted data for previously deleted instances.Alex Meade2012-07-231-0/+67
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | These migrations undelete metadata, system metadata, and block device mappings that were deleted when instances were deleted before the change: 0598db0a6c66eb2b6d2e0e3f0b3b2b921ebe8621 Before that change; metadata, system metadata, and block device mappings were deleted when and instance was deleted, this is no longer the behavior. These migrations undelete all of the data that was deleted within two seconds of the instance being deleted in order for old deleted instances to be consistent with the new behavior. Fixes bug 1022592 Change-Id: I36c66913abcb231f23e409f1f71c00ef0f6e69e5
* | | | Merge "fixes for nova-manage not returning a full list of fixed IPs"Jenkins2012-07-231-1/+1
|\ \ \ \
| * | | | fixes for nova-manage not returning a full list of fixed IPsTong Li2012-07-201-1/+1
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fix correct the problem reported in bug 930193 nova-manage does not return a full list of fixed IPs. The reason is that the call to fixed_ip['instance_id'] can throw exception when fixed_ip does not have instance_id as its member. The changes in this fix corrects the problem. rebase to resolve the automatic merge issue. Change-Id: Ic15a0a0f43f53b82f9d88a03803c96b44a8ddcb9
* | | | Merge "Fix EC2 CreateImage no_reboot logic"Jenkins2012-07-232-3/+26
|\ \ \ \
| * | | | Fix EC2 CreateImage no_reboot logicEoghan Glynn2012-07-172-3/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously the call to restart the imaged instance was incorrect. This is now fixed and unit tested. Change-Id: Ifde193e084ecdbc37a57150f28d833ae276f49c6
* | | | | Merge "Reject EC2 CreateImage for instance-store"Jenkins2012-07-233-6/+68
|\| | | | | |/ / / |/| | |
| * | | Reject EC2 CreateImage for instance-storeEoghan Glynn2012-07-173-6/+68
| | | | | | | | | | | | | | | | | | | | | | | | | | | | For fidelity with AWS, we explicitly do not support the EC2 CreateImage API for the analogue of instance-store-backed instances. Change-Id: I266a021ce53dbb1f521201f1cdc614b08c31bf8d
* | | | Merge "Remove unnecessary use of with_lockmode"Jenkins2012-07-201-1/+1
|\ \ \ \ | |_|/ / |/| | |
| * | | Remove unnecessary use of with_lockmodeJohannes Erdfelt2012-07-201-1/+1
| | |/ | |/| | | | | | | | | | | | | | | | Using with_lockmode outside of a transaction is useless. It's also unnecessary when code won't ever make any updates to the records. Change-Id: Ib58b035cc5803e49654570b451ad381bebdd300c
* | | Merge "Make flavorextradata ignore deleted flavors."Jenkins2012-07-201-1/+1
|\ \ \
| * | | Make flavorextradata ignore deleted flavors.David Besen2012-07-181-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | In _get_flavor_refs, don't retrieve inactive flavors. Fixes bug 1026210. Change-Id: I7c5d7848d2d1326d01770195e1c8a754bb673a7a
* | | | Merge "set correct SELinux context for injected ssh keys"Jenkins2012-07-202-0/+37
|\ \ \ \
| * | | | set correct SELinux context for injected ssh keysPádraig Brady2012-07-202-0/+37
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instruct guests to ensure at boot, that the correct SELinux context is set for /root/.ssh/. This will cater for keys injected by nova from hosts without SELinux (enabled) or using libguestfs which currently doesn't support setting extended attributes. Suggested-by: David Naori <dnaori@redhat.com> Change-Id: Ibf3869e3ee477e91623e0c030838c1ec8a6128a6
* | | | | Merge "Improve VDI chain logging."Jenkins2012-07-201-28/+11
|\ \ \ \ \
| * | | | | Improve VDI chain logging.Rick Harris2012-07-201-28/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently we're logging the parent's OpaqueRef which is rather useless. This patches uses the much more useful UUID. Some small cleanups are included as well. Change-Id: I6a9dd712dbac1d1b1aaba65da2b7c399b606e934
* | | | | | Merge "Support requested_networks with quantum v2."Jenkins2012-07-203-6/+36
|\ \ \ \ \ \ | |_|_|_|/ / |/| | | | |
| * | | | | Support requested_networks with quantum v2.Salvatore Orlando2012-07-143-6/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow Quantum to attach instance only to selected networks, which can be specified, for instance, via the --nic option on the nova CLI. Fixes bug 1019899 Change-Id: I099056387edea9e3646cd4e329c76b674e03baa2
* | | | | | Merge "Fixes bare-metal spawn error."Jenkins2012-07-201-18/+4
|\ \ \ \ \ \
| * | | | | | Fixes bare-metal spawn error.Mikyung Kang2012-07-171-18/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Implements blueprint Bare-metal provisioning. Fixes bug 1025942. Change-Id: I55d16747d117de728b02ac938083ef8a79eaaca7