In oslo.limit 2.6.0 service endpoint discovery was added, provided by
three new config options:
[oslo_limit]
endpoint_service_type = ...
endpoint_service_name = ...
endpoint_region_name = ...
We can use the same config options if they are present to lookup the
service ID and region ID we need when calling the
GET /registered_limits API as part of the resource limit enforcement
strategy. This way, the user will not have to configure endpoint_id.
This will look for [oslo.limit]endpoint_id first and if it is not set,
it will do the discovery.
Closes-Bug: #1931875
Change-Id: Ida14303115e00a1460e6bef4b6d25fc68f343a4e
These are *super* annoying (and useless to boot, since there is nothing
we can do about them in the near term). Shut them ⬇️⬇️⬇️ down ⬇️⬇️⬇️.
Change-Id: I469dafa243b95749b34503c1f3e905d9d8c780d4
Signed-off-by: Stephen Finucane <stephenfin@redhat.com>
The target goal of these series of patch is to enable VFIO devices
migration with kernel variant drivers.
Partially-Implements: blueprint migrate-vfio-devices-using-kernel-variant-drivers
Change-Id: I23af0d36448e9b659f6383d602d9dfa0e2798e60
Adds the field ``pci_dev_map_src_dst`` to ``LiveMigrateData`` object
to include the PCI device mapping between the source and destination
device addresses.
The target goal of these series of patch is to enable VFIO devices
migration with kernel variant drivers.
Partially-Implements: blueprint migrate-vfio-devices-using-kernel-variant-drivers
Depends-On: https://review.opendev.org/c/openstack/nova/+/938405
Change-Id: I1c8e5709640abedb332d96c2c1f5268fa239d9a1
The target goal of these series of patch is to enable VFIO devices
with kernel variant drivers.
Implements: blueprint enable-vfio-devices-with-kernel-variant-drivers
Change-Id: I7949ba6da8b6257865d8e9e48bf3feabc10bdf17
microversion 2.98 changed the GET and GET list details
API response to expand the image field
- https://review.opendev.org/c/openstack/nova/+/939649
but when we change the views/servers.py then update server
and rebuild server API response also gets updated.
- https://review.opendev.org/c/openstack/nova/+/939649/12/nova/api/openstack/compute/views/servers.py#595
The microversion change actually changed the response for
the update and rebuild API but did not update the documentation
about this change. Neither added test for rebuild and update
which verify the response change.
Since microversion 2.75, we have made all these 4 APIs response
consistent.
Implements: blueprint image-properties-in-server-show
Change-Id: Ifc2edc59c601391a31fd8953c742969e5b368767
microversion 2.96 changed the GET and GET list details
APi response to add a new field 'pinned_availability_zone'
- https://review.opendev.org/c/openstack/nova/+/904568
but when we change the views/servers.py then update server
and rebuild server API response also gets updated.
- https://review.opendev.org/c/openstack/nova/+/904568/17/nova/api/openstack/compute/views/servers.py#309
The microversion change actually changed the response for
the update and rebuild API but did not update the documentation
about this change. Neither added test for rebuild and update
which verify the response change.
Since microversion 2.75, we have made all these 4 APIs response
consistent.
Change-Id: I31fbb3c5e689302713bfb7a254c5d0cc12de2dc4
Currently, even when [libvirt] volume_use_multipath is set to True,
volume attachment silently falls back to single path if multipathd is
not running in the host. This sometimes prevents operators from
noticing the misconfiguration, until they face any issue caused by
missing redundancy.
Introduce the new [libvirt] volume_enforce_multipath option, which
makes the attachment process fail if multipathd is not running.
A similar parameter (enforce_multipath_for_image_xfer) was already
implemented in cinder and this change follows how the parameter is
implemented there.
Also add the check in init phase to detect lack of mulitipath daemon
during initializing driver.
Min version of os-brick has to be bumped due to the interface change
made by 8d919696a9f1b1361f00aac7032647b5e1656082 .
Implements: blueprint enforce-multipath
Change-Id: I828de70ca7b343a4562ace4049d2b3857dbf900a
Nova currently lacks a straightforward way to expose scheduler hints
associated with a server. This change extends existing Nova's
API to allow users to retrieve this information when it is available.
It adds a new API microversion to support returning the associated
scheduler_hints in ``GET /servers/{server_id}``, `GET /servers/detail``,
``PUT /servers/{server_id}`` and ``POST /server/{server_id}/action``
(rebuild) responses.
Implements: blueprint show-scheduler-hints-in-server-details
Change-Id: If0070cfbc218e0515bf3b58fc8c5067ae22c7072
The target goal of these series of patch is to enable VFIO devices
with kernel variant drivers.
Implements: blueprint enable-vfio-devices-with-kernel-variant-drivers
Change-Id: Iea3efa29b6b89639b21d3e701f8710d0043d00c7
This patch adds a new console type, "spice-direct", which provides
the connection information required to talk the native SPICE protocol
directly to qemu on the hypervisor. This is intended to be fronted
by a proxy which will handle authentication separately.
A new microversion is introduced which adds the type "spice-direct"
to the existing "spice" protocol.
An example request:
POST /servers/<uuid>/remote-consoles
{
"remote_console": {
"protocol": "spice",
"type": "spice-direct"
}
}
An example response:
{
"remote_console": {
"protocol": "spice",
"type": "spice-direct",
"url": "http://localhost:13200/nova?token=XXX";
}
}
This token can then be used to lookup connection details for the
console using a request like this:
GET /os-console-auth-tokens/<consoletoken>
Which returns something like this:
{
"console": {
"instance_uuid": <uuid>,
"host": <hypervisor>,
"port": <a TCP port number>,
"tls_port": <another TCP port number>,
"internal_access_path": null
}
}
APIImpact
Change-Id: I1e701cbabc0e2c435685e31465159eec09e3b1a0
This patch makes just the schema changes required for the
implementation of SPICE direct consoles, as requested during
review. See change I1e701cbabc0e2c435685e31465159eec09e3b1a0
for the related feature implementation.
APIImpact
Change-Id: I838ad1a8a74f47544226d3da0e7f1a2f5585b7cb
This patch changes just the objects required for the implementation
of SPICE direct consoles, as requested during review. See change
I1e701cbabc0e2c435685e31465159eec09e3b1a0 for the related feature
implementation.
APIImpact
Change-Id: I8ec513fd818362d2f041c679a84e900952af61a2
This a partial revert of change
I60d6f04d374e9ede5895a43b7a75e955b0fea3c5 which added tpool.Proxy
wrapping to the listDevices() and listAllDevices() methods.
The regression was caught during downstream testing with vGPUs and the
update_available_resource() periodic task was failing with:
TypeError: virNodeDeviceLookupByName() argument 2 must be str or
None, not Proxy
It turns out that while the listAllDevices() method returns a list of
virNodeDevice objects [1], the listDevices() method returns a list of
string names [2] and is generated from the corresponding function in C
[3].
The error was not caught by unit or functional testing because those
test environments intentionally do not import the libvirt Python
module -- so mocked code in the LibvirtFixture runs instead. Also, the
update_available_resource() method has a 'except Exception:' at the end
which logs an error but does not re-raise. So it would not cause a
functional test to fail.
This reverts the change that caused the regression, updates potentially
confusing docstrings, adds type annotations to the methods that use
listDevices(), and moves the nodeDeviceLookupByName type checking into
the LibvirtFixture.
Closes-Bug: #2098892
[1] https://github.com/libvirt/libvirt-python/blob/408815a/libvirt-override-virConnect.py#L520-L524
[2] https://github.com/libvirt/libvirt-python/blob/408815a/libvirt-override-api.xml#L448-L453
[3] https://libvirt.org/html/libvirt-libvirt-nodedev.html#virNodeListDevices
Change-Id: Ib5befdd3c13367daa208ff969f66cba693ae2c76
Change I60d6f04d374e9ede5895a43b7a75e955b0fea3c5 added tpool.Proxy
wrapping to the listDevices() and listAllDevices() methods but
introduced a regression for listDevices() that led to an error in
update_available_resource():
TypeError: virNodeDeviceLookupByName() argument 2 must be str or
None, not Proxy
The error was not caught by unit or functional testing because those
test environments intentionally do not import the libvirt Python
module -- so mocked code in the LibvirtFixture runs instead. Also, the
update_available_resource() method has a 'except Exception:' at the end
which logs an error but does not re-raise. So it would not cause a
functional test to fail.
This adds a functional test to reproduce the bug and adds a keyword arg
to the test _run_periodics() method to specify whether it should raise
an exception if an error is logged.
Related-Bug: #2098892
Change-Id: I3a3dda57f2181b24bd6589ac7bb8160014ab2396
This weigher will check how many instances in the host have the image
properties that are requested and will prefer by default to pack
instances with the same properties.
Implements blueprint: image-metadata-props-weigher
Change-Id: I3bfed44bd089c6b226d13c3ac4a0003411737cbd
In centos/rhel 9 qemu supprot for the qxl model was removed
with the removal of spice support.
In ubuntu 24.04 qemu support for qxl and spice supprot
has now also been removed.
debian 12 bookworm still support spice in there qemu package.
When we updated the default video model to virtio for x86 we
left a config driven special case for spice to default to qxl
since that nolonger works on centos or ubuntu based distos that
default is not helpful so this change removes the special case
making virtio the default for x86 regardless of the console used.
This change also updates the nova-ovs-hybrid-plug to test with spice
so that we have at least one job using it. to enable that the job is
moved to debian.
Closes-Bug: #2097529
Change-Id: I265ad2ced3729bed41bf53c58dcebadb775ce1f7
This adds a non-remotable method to InstanceList which will batch-
fill system_metadata for all the instances in the list that are
missing it in as efficient of a manner as possible. This does not
require an object bump because no remotable methods or fields are
changed.
Related to blueprint image-metadata-props-weigher
Change-Id: Icc47de2b677b3d212a7f6faa61a85ea9bff9f412
melanie witt