| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
In _destroy_vdis() in the xenapi vmops.py, a list of VDIs to not
destroy is collected. If one of the block device mappings does
not have a corresponding VDI, however, _destroy_vdis() fails
(KeyError('vdi_uuid') is raised). This interferes with
instance deletion. The problem is corrected by skipping mappings
for which the connection_info data does not contain 'vdi_uuid'.
Fixes bug 1030143.
Change-Id: I5125d557ab799fcf572f665523cd07a893ad3b90
|
| |\ \ \ \ \ \ |
|
| | | |_|/ / /
| |/| | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The preferred form of calling the DB API is to use the instance UUID
instead of the older instance id. Update that and use the dict form
to be more compatible with blueprint no-db-messaging.
Change-Id: Ib1a06b5385ec627f3daad5bafb3fa1be956a64e2
|
| |\ \ \ \ \ \ |
|
| | |/ / / / /
| | | | | |
| | | | | |
| | | | | | |
Change-Id: Ibd28e8de19eb6dc696ab67d2f58b4a14eb9f034f
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This fixes bug 1030106. Basically we were not counting disabled hosts
in the total host count. Also I have refactored the get_audit_task_logs
method, it's really part of the extension (only the extension uses it),
so I've moved it out of nova.compute.utils into the extension itself.
(the tests that exercise it already lived in the extension's tests anyway.)
Change-Id: Iaf42d887b824ba0cbf6ab0ed143a4c01b37b0cb1
|
| | |/ / / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The XenAPI driver would use _rescue to note when an instance was in
the process of being rescued. The existance of this attribute would
modify the 'name' property to append '-rescue' to the generated name.
Among the problems with it, it wasn't saved anywhere (with resulting
code to set it to False being unnecessary) and it led to "magic" code
(see instance.name) that was confusing without understanding the hack.
Change-Id: I1a91e4efdf69d4c1fba07658419f7bb30b733b5a
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Fixes bug 1030073
With instances being passed as dicts via the RPC API, uses of the instance
object should use dict style key lookups instead of attribute style
lookups.
Change-Id: Iefc0e403d440aada68b259ded457166ad10699fd
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Xenstore only allows certain characters in key names. Change disallowed
characters as well as '/' to '_'.
Fixes bug 1029773
Change-Id: I04055bfbe662f3f3e9d90336d03670aa5468e780
|
| |\| | | | |
|
| | | |_|/
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Fixes bug 1029774
system_metadata should not be exposed to users.
This reverts part of commit 82afe7ad5 / ChangeID I88a59f1b783
Change-Id: Id454dbcbafef861db3f88620a56d8dd2ccd30e1d
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Now sends a notification of other changes in the instance besides state
changes. Adds access_ip in notifications as well.
fixes bug 1028559
Change-Id: I664e6b858eda17dd732927f2c9b259212084a8fa
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Adding config drive as an extension allows users to turn it off in case they
do not want to use it. Config drive was not intended to be a part of the core
functionality.
Fixes bug 833331
Change-Id: Icf4b5a21d8cc9df0f140a1a87908ff48c447004b
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This reverts commit bd30eb36bbf2c5164ac47256355c543f6b77e064.
From the original commit message:
" This is an interim solution to provide this functionality until
it is replaced by more general HostAggregate functionality later
in Folsom."
Now that general host aggregates is in, this is ready to be removed.
Partially implements blueprint general-host-aggregates
Change-Id: Ic5f418a5d15117888198c79c3a32cec8f19d1a88
|
| |\ \ \ \ \ \ |
|
| | | |_|/ / /
| |/| | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Fixes bug 1021352.
In network/manager/deallocate_fixed_ip the call
to mark the IP as no longer allocated occurs before
the call to update security group rules. This means
that if an error occurs in the security group
processing, or if for some reason it is very slow
there is a risk that that the address is reused by
another tenant before the rules relating to this address
have been fully revoked.
Moving the db call to after the call to trigger the
security group refresh would avoid the issue when an
error occurs (the fixed_ip should not be released in
this case).
Change-Id: Iaba1af5c9a17fbbb82e42522b1060773de61550a
|
| |\ \ \ \ \ \ |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the pre_live_migration method of the compute
rpc API to take a full instance over rpc instead of just
the instance UUID. This cuts down on database access needed
by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: Ib628d375254fc1539cd79ceaaad83c1c326863b4
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the power_on_instance and start_instance methods of the compute
rpc API to take a full instance over rpc instead of just
the instance UUID. This cuts down on database access needed
by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: I5a95e72c6441f47f014dbcd52f346e8d2e94069c
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the power_off_instance and stop_instance methods of the compute
rpc API to take a full instance over rpc instead of just
the instance UUID. This cuts down on database access needed
by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: I641d5e7941c7de430c0ca3c9b961f967d947d963
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
In some refactoring of the decorator modifications I made for
eebc64f949ccb2acb7462efc18f538f1827985af, I ended up breaking the case
where an instance_uuid was passed to one of the methods that is supposed
to be able to accept either an instance or instance_uuid. Oops. :-(
This patch makes them work again, and adds one unit test that
verifies that it works.
These decorators are really ugly right now, but keep in mind that the
2-version stuff is temporary. As soon as this patch series is finished
(changing methods to take an instance), these decorators will get
simplified.
Part of blueprint no-db-messaging.
Change-Id: Ic1b1269d90e147035c4eb7b4bd47de81eff3ffdc
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the post_live_migration_at_destination method of the compute
rpc API to take a full instance over rpc instead of just
the instance UUID. This cuts down on database access needed
by nova-compute.
The method got moved in rpcapi.py, but that was just to restore
alphabetical order.
Part of blueprint no-db-messaging.
Change-Id: Ib36cb065f838f644e97b6e82b39409737df15558
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the inject_network_info method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: I1bcf55e2819be321c3ea0a64fe136cd993a2443b
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the inject_file method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: Ife38a87cb8a597b40b2b87d9a6a8c7e81d4439c4
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the get_vnc_console method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: I4443e65fd6a9aafdad21fda63ef04a03428288a9
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
This method is no longer used in any current code, so remove it from the
client side code for the rpcapi. The server-side code remains (the
compute manager) for backwards compatibility with older versions.
This method was previously used by the scheduler, but was removed by
some refactoring done in cac332c39645286a11c009094a86f62d02752183.
Change-Id: Icb11001ac0880ca8df67b4dd9f0fe1eaca7fa3bb
|
| |\| | | | | | |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Change the get_diagnostics method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: Ic4db237849230899ad5f2f0780d4c237895a6631
|
| |\| | | | | |
| |/ / / / /
|/| | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change the finish_revert_resize method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: Ibe2f8b990cc8f2a3d05e8ba8216a1e74cbcfb67b
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This fixes bug 1029721 which was causing instance spawn failures.
Change-Id: Iab7fdb063874972e0bcd91c51bfd0e5b49e8023c
|
| |\ \ \ \ \ \
| |/ / / / /
|/| | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Commit 31336b35 removed architecture from the instance base_options
passed to _schedule_run_instance(). The architecture specified in
the image is needed by a subsequent patch to the ComputeFilter which
filters compute hosts against the triple (architecture, hypervisor_type,
vm_mode).
Change-Id: I840356979e51de3a6d941fe4ae9610e488383a44
|
| |\ \ \ \ \ \ |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
blueprint update-flavor-key-value
Enhance nova-manage so that it can add, delete and list key/value pairs for
flavors extra specs.
Change-Id: I0665bec6aecd7f7a1d8e845fd930f4bd5a841522
Signed-off-by: Yunhong, Jiang <yunhong.jiang@intel.com>
|
| |\ \ \ \ \ \ \ |
|
| | | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
Change the finish_resize method of the compute rpc API to take
a full instance over rpc instead of just the instance UUID.
This cuts down on database access needed by nova-compute.
Part of blueprint no-db-messaging.
Change-Id: I2b0c7cbd2ba9b6243c2c50e69e894d192e230491
|
| |\ \ \ \ \ \ \ \ |
|