The switch to using glance from the SDK accidentally used get_image
directly during a server rebuild, when it should have used find_image to
match existing functionality.
Bug introduced in: I36f292fb70c98f6e558f58be55d533d979c47ca7
Change-Id: I2005bd40a1bd6719670c7f7854316b4f9801b140
Story: 2007620
Task: 39643
This is a work to switch OSC from using glanceclient to OpenStackSDK.
With this change only v2 is using OpenStackSDK. V1 is still using
glanceclient and will be switched in a separate change.
Remove the direct depend on keystoneauth- let that flow through
openstacksdk.
Depends-on: https://review.opendev.org/#/c/698972
Change-Id: I36f292fb70c98f6e558f58be55d533d979c47ca7
Does what it says on the tin. This action was added to osc-lib in change
If73cab759fa09bddf1ff519923c5972c3b2052b1.
Change-Id: I51efaa096bb26e297d99634c5d9cca34c0919074
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
The '--hint' option for 'server create' expects a key-value pair like so:
openstack server create --hint group=245e1dfe-2d0e-4139-80a9-fce124948896 ...
However, the command doesn't complain if this isn't the case, meaning
typos like the below aren't indicated to the user:
openstack server create --hint 245e1dfe-2d0e-4139-80a9-fce124948896
Due to how we'd implemented this here, this ultimately results in us
POSTing the following as part of the body to 'os-servers':
{
...
"OS-SCH-HNT:scheduler_hints": {
"245e1dfe-2d0e-4139-80a9-fce124948896": null
}
...
}
Which is unfortunately allowed and ignored by nova due to the use of
'additionalProperties' in the schema [1]
Do what we do for loads of other options and explicitly fail on invalid
values. This involves adding a new argparse action since none of those
defined in osc-lib work for us. This is included here to ease
backporting of the fix but will be moved to osc-lib in a future patch.
[1] https://github.com/openstack/nova/blob/19.0.0/nova/api/openstack/compute/schemas/servers.py#L142-L146
Change-Id: I9e96d2978912c8dfeadae4a782c481a17cd7e348
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Story: #2006628
Task: #36840
Related-Bug: #1845322
1. As mentioned in [1], we should avoid using six.iteritems to achieve
iterators. We can use dict.items instead, as it will return iterators
in PY3 as well. And dict.items/keys will more readable.
2. In py2, the performance about list should be negligible,
see the link [2].
[1] https://wiki.openstack.org/wiki/Python3
[2] http://lists.openstack.org/pipermail/openstack-dev/2015-June/066391.html
Co-Authored-By: Akihiro Motoki <amotoki@gmail.com>
Change-Id: I4b9edb326444264c0f6c4ad281acaac356a07e85
Implements: blueprint replace-iteritems-with-items
This patch removes using the "name" option for a marker when
--deleted is also used. The find_resource() function
that is being called does not correctly handle using the marker
as the "name" in the search when also using deleted=True.
One simple way to fix this is force the marker to only be an ID
when --deleted is used. This is how the nova client works.
Using the --deleted option is available to users with the admin
role by default. If you're an admin listing --deleted servers
with a marker by name, find_resource() is going to fail to find
it since it doesn't apply the --deleted filter to find_resource().
The find_resource() function is trying to find the marker server
by name if it's not found by id, and to find it by name it's
listing servers with the given marker as the name, but not
applying the --deleted filter so it doesn't get back any results.
In the story it was suggested modifying find_resource to include
the deleted query param when it's specified on the command line but
that didn't work because it still results in something like this:
http://192.168.1.123/compute/v2.1/servers?deleted=True&name=4cecd49f-bc25-4a7e-826e-4aea6f9267d9
It seems like there are bugs in find_resource().
Restricting the marker to be the server ID when listing deleted servers
is probably OK since if you're using --deleted you're an admin and you could
be listing across all projects and if you're filtering by a server across all
projects anyway (not that you have to, I'm just saying if you are), or even
showing a server in another project, you have to do it by id rather than name
because find_resource() won't find the server in another project by name, only ID.
story: 2006761
Task: 37258
Change-Id: Ib878982b1d469212ca3483dcfaf407a8e1d2b417
Added ``--disable-delete-on-termination`` and
``--enable-delete-on-termination`` options to the
``openstack server add volume`` command that enables users to mark
whether to delete the attached volume when the server is destroyed.
Depends-On: https://review.opendev.org/#/c/681267/
Part of blueprint support-delete-on-termination-in-server-attach-volume
Change-Id: I6b5cd54b82a1135335a71b9768a1a2c2012f755b
While cold migration and resize are essentially the same operation under
the hood, meaning one could use the 'openstack server resize confirm'
and 'openstack server resize revert' commands instead, there is no
reason the operator needs to know this. Add these flags as syntactic
sugar to help simplify operators lives.
The help texts for both the 'openstack server resize' and 'openstack
server migrate' commands are updated to clarify the relationship between
the two operations.
Change-Id: I0cb6304c794bffaec785add9f7b8cf53ab28cacd
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
This patch adds a new parameter ``--availability-zone`` to
``openstack server unshelve`` command. This can help users to specify
an ``availability_zone`` to unshelve a shelve offloaded server from
2.77 microversion.
Depends-On: https://review.opendev.org/679295
Implements: blueprint support-specifying-az-when-restore-shelved-server
Change-Id: Ia431e27c2a17fe16466707cc362532860ecf22df
Pick up newer versions of this library. Thankfully no serious changes
are needed.
Change-Id: I69e523844529fc1c8aa0c1ce764182dbe29cfeb6
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
This adds a --boot-from-volume option to the server create
command which is used with the --image or --image-property
option and will create a volume-backed server from the
specified image with the specified size. Similar to the
--volume option, the created root volume will not be deleted
when the server is deleted. The --boot-from-volume option
is not allowed with the --volume option since they both create
a block device mapping with boot_index=0.
Change-Id: I88c590361cb232c1df7b5bb010dcea307080d34c
Story: 2006302
Task: 36017
The <id> portion of a --block-device-mapping value can be
the resource name or id since the code uses the appropriate
type-specific find_resource utility to lookup the resource
based on the value given.
This change simply fixes the description of <id> to mention
it's name or id rather than just "UUID". My guess is the
description was originally copied from novaclient where id
must be an id since name resolution does not happen in novaclient.
Change-Id: I567f6f6efb3a3b6d387133d21aa81354b2d753bc
The --block-device-mapping option on the server create
command currently only supports booting from volume and
volume snapshot. A common boot-from-volume scenario is
providing an image and letting nova orchestrate the
creation of the image-backed volume and attaching it to
the server.
This adds support for type=image in the --block-device-mapping
option. The volume size is required in this case. Note that
the CLI currently says if type=snapshot that size is also required
but that's technically not true. When booting from a volume
snapshot, the compute API will use the size of the volume snapshot
to create the volume if an explicit size is not provided. For the
purposes of this patch, we need the size anyway for the image
being the block device mapping source type.
Change-Id: I57b3c261d8309f7b9f62a3e91612bce592a887a3
Story: 2006302
Task: 36016
These are currently exposed as flags on the 'openstack server resize'
command but they are in fact operation and should be exposed as commands
in their own right.
The old flag-based variants are deprecated for removal in 4.0.
Change-Id: I733796d3bda6c3755a3d3548bbe695abb474a6a0
Signed-off-by: Stephen Finucane <sfinucan@redhat.com>
Adds the --host and --hypervisor-hostname options to
``openstack server create`` CLI.
Depends-On: https://review.opendev.org/670558
Change-Id: If188c3d96fa506dbe62ef256418f2f9bca1520c2
Blueprint: add-host-and-hypervisor-hostname-flag-to-create-server
When the --name-lookup-one-by-one option passed to the 'server list'
command, the image and flavor names will be looked up for each
server being listed instead of fetching all image/flavor names.
The current code assumes all servers have an image attribute, but
servers booted from volumes have no image, so the following error is
raised when listing BFV servers with --name-lookup-one-by-one:
AttributeError: ('unicode'|'str') object has no attribute 'get'
The error occurs when the code attempts server.image.get('id').
This fixes the --name-lookup-one-by-one code not to assume an image
for a server. The unit tests for 'server list' have also been
robustified to feature one BFV server to enhance our test coverage.
Story: #2006063
Task: #34777
Change-Id: I312c971346c7ded93f6fcaa515098554b8580295
This patch adds a new parameter ``--reason`` to ``openstack server lock``
command and ``--locked``, ``unlocked`` filtering parameters to
``openstack server list`` command. This can help users to provide a
reason when locking the server and to filter instances based on their
locked value from 2.73 microversion.
Implements blueprint add-locked-reason
Depends-On: https://review.opendev.org/#/c/661785/
Change-Id: Ib2714f98b24d47e570da8a6c231e765acd2ff595
Currently, doing a cold migration while specifying a target host is not
possible however nova api supports it since version 2.56.
This patch allows passing "--host" when doing a cold migration.
It runs normally if --os-compute-api-version is 2.56 or greater and
returns an error otherwise.
Change-Id: I960109008096ce8bb4e4c8ca6ffb22c33aacd995
Story: 2003325
Task: 24359
The compute API 2.20 microversion allows attaching and detaching
a volume to/from a server with status SHELVED or SHELVED_OFFLOADED.
For OSC this just means the user has to specify the appropriate
minimum microversion to make that work, so this change mentions
that in the "server add volume" and "server remove volume" command
description.
Related to nova blueprint volume-ops-when-shelved
Change-Id: I4824175e5d9e124e3bd9e9a8fd5a89277efc6cff
Per the discussion at the Train Forum [1] this deprecates
the problematic --live option on the server migrate command
which, depending on the compute API version used, forcefully
bypasses the scheduler and also does not allow you to live
migrate a server and let the scheduler pick a host.
The --live option is replaced here with two new options:
* --live-migration: this simply tells the command you want to
perform a live rather than cold migration; if specified with
--live the --live-migration option takes priority.
* --host: when specified, this will request a target host for
the live migration and will be validated by the scheduler;
if not specified, the scheduler will pick a host. This option
is mutually exclusive with --live.
We can build on the --host option by supporting cold migrations
with a specified host when using compute API version 2.56 or
greater but that will come in a separate change.
If the --live option is ever used we log a warning.
Note there are several related changes for this issue:
- https://review.openstack.org/#/c/628334/
- https://review.openstack.org/#/c/626949/
- https://review.openstack.org/#/c/627801/
- https://review.openstack.org/#/c/589012/
- https://review.openstack.org/#/c/460059/
This change allows us to deprecate the --live option and provide
a replacement which is backward compatible without having to use
something potentially error-prone like nargs='?'.
Closes-Bug: #1411190
[1] https://etherpad.openstack.org/p/DEN-osc-compute-api-gaps
Change-Id: I95d3d588e4abeb6848bdccf6915f7b5da40b5d4f
This patch adds functionality to configure server's description with:
1 server create
2 server set
3 server unset
4 server rebuild
Change-Id: Ic06d97b29e51828b29d7ac5172645c288e4ada9e
Story: 2002005
Task: 19640
There is no indication to the user in the command
help that they have to use 2.17 or greater [1] to run
the "openstack server dump create" command. This
mentions that requirement in the help of the command.
[1] https://developer.openstack.org/api-ref/compute/#trigger-crash-dump-in-server
Change-Id: I02c06e10a26eb38ddecb70f970cfcbfad962201c
This patch explicitly points out the change needed while
forming the detailed lists for servers. In those cases
where the server response for ``openstack server list``
has the flavor and image keys missing for the instances
in the down cell, the servers will be skipped from being
processed.
Depends-On: https://review.openstack.org/591657/
Related to blueprint handling-down-cell
Change-Id: Ibcfe9febdc45db1cb86c6e88f65976feceb01c02
AddFloatingIP refers to an old nova proxy API to neutron that was
deprecated in nova. The neutron API for floating IP associate requires
a port to be specified. Currently, the code is selecting the first port
if the server has multiple ports. But, an attempt to associate the
first port with a floating IP can fail if the first port is not on a
network that is attached to an external gateway.
In order to make the command work better for users who have a server
with multiple ports, we can:
1. Select the port corresponding to the fixed_ip_address, if one was
specified
2. Try to associate the floating IP with each port until one of the
attempts succeeds, else re-raise the last exception.
(404 ExternalGatewayForFloatingIPNotFound from neutron)
This also fixes incorrect FakeFloatingIP attributes that were being set
in the TestServerAddFloatingIPNetwork unit tests, which were causing
the tests to use None as parsed args for ip-address and
--fixed-ip-address and thus bypassing code in the
'if parsed_args.fixed_ip_address:' block.
Task: 27800
Story: 2004263
Change-Id: I11fbcebf6b00f12a030b000c84dcf1d6b5e86250
usually in a big cloud there are many images and flavors,
while each given project might use only some of those.
This patch introduces '--name-lookup-one-by-one' argument to
server list command (mutually exclusive with '--no-name-lookup')
When provided (or either '--image' or '--flavor' is specified) to the
`server list` command, name resolving for
corresponding entity is now using targeted GET commands instead of
full entities list.
In some situations this can significantly speedup the execution of the
`server list` command by reducing the number of API requests performed.
Change-Id: I59cbf3f75c55e5d3747654edcc9be86ad954cf40
Story: #2002039
Task: #19682
Add '--property' option to the 'server rebuild' command, to provide
the ability to specify properties of the rebuilt instance.
This is equivalent to the '--meta' option of the compute's
'nova rebuild' command.
Change-Id: I25ea6622e970416090109316e1e28fab8b0b3f07
Story: #2003979
Task: #26922
In API microversion 2.25 Nova removed the disk_over_commit parameter
to the live migration server action. This patch makes sure that we
don't include it in our request if we're running with 2.25 or higher.
Story: #2002963
Task: #22966
Change-Id: I1bbdd33be96d82422a05982508e370237c3560f3
save (potentially many) HTTP calls to Glance API for image list
and a call to Nova API for flavor list when the server list
actually returned no servers.
Change-Id: I93a56138c50b82fb4dce67a2f788107f71c5f423
Story: #2002039
Task: #19681
Compute API version 2.47 embeds the server's internal
flavor in the response. The original flavor id is not
preserved since it could have changed if the flavor
was deleted and re-created after the server was created,
which was the dreaded Horizon "Edit Flavor" issue.
So the flavor dict in the server response is a dict of
information about the flavor representing the server
"right now" excluding the id. The original flavor name
is shown though along with the ram/disk/vcpu etc
information.
The server list command has a similar issue which
will be fixed in a follow up change.
Change-Id: I1a92999758006d02567c542b6be8902a049899cc
Task: 13864
Story: 1751104
When showing a server or doing a rebuild, we already have
the latest version of the server so _prep_server_detail
getting the server again is an unnecessary performance hit.
ShowServer is pretty obvious here. For RebuildServer, the
compute API actually refreshes the server before returning
it in the response, so the client already gets the latest
when the rebuild call returns.
The only other usage of _prep_server_detail that does require
a refresh is CreateServer since the POST /servers response is
a minimal version of the server object.
This adds a new refresh kwarg, backward compatible by default,
to _prep_server_detail but changes ShowServer and RebuildServer
to no longer refresh.
Change-Id: Ib1c9c424ed1cafc2dfd8be90af8de8a774bdfbf0
There is a to_dict() method on the Server object from
python-novaclient which makes a deepcopy of the internal
Server._info - use this instead of accessing the _info
attribute directly.
Also, while in here, fixed a typo in _prep_server_detail.
Change-Id: I679b4489c815f8a54368ef6b23b9f77e75b4d0bc
