This is a mechanically generated change to replace openstack.org
git:// URLs with https:// equivalents.
This is in aid of a planned future move of the git hosting
infrastructure to a self-hosted instance of gitea (https://gitea.io),
which does not support the git wire protocol at this stage.
This update should result in no functional change.
For more information see the thread at
http://lists.openstack.org/pipermail/openstack-discuss/2019-March/003825.html
Change-Id: Ia9081a1f0249a7a8b863093faa0f0b05f27cbb85
server filter/sort spec (add-whitelist-for-server-list-filter-sort-parameters.rst)
is implemented in ocata but there are 'all_tenants' behavior
and policy changes mentioned in this spec which actually did not get
implemented.
Those proposed changes were rediscussed in API meeting [1] to implement in
a different way [2] which again not completed yet.
There was no consensus on direction that time and need more broad discussion
to make 'all_tenants' bits as better UX.
But anyways we should cleanup the implemented spec with things
not yet done. Otherwise people might get wrong impressions about 'all_tenants'
changes.
This commit does the below change to server filter/sort spec:
1. Some valid params which are implemented in schema but missing in
spec's whitelist:
- Add 'created_at' and 'changes-since' in filter params list
- Add 'created_at' and 'updated_at' in sort param list
2. Remove the 'all_tenants' proposed changes which are not implemented yet.
We will further discuss/plan the 'all_tenants' proposal as fresh spec.
.. 1 http://eavesdrop.openstack.org/meetings/nova_api/2017/nova_api.2017-01-25-13.00.log.html
.. 2 https://review.openstack.org/#/c/426128/https://review.openstack.org/#/c/425533/
Change-Id: Ibdcc5cae3b73c0aa0d3e0a785c827682df0483ca
In commit a67d39ba, we removed templates from both the approved and
implemented folders of 'specs/{release}'. However, git tracks files, not
folders, meaning the 'implemented' folders for both Ocata and Pike,
which are currently empty, are no longer tracked and the glob patterns
used in 'doc/source/specs/{release}/index.rst cause warnings.
These templates are re-added to fix the immediate issue. However, we
still don't want to keep these around once the release has been made. As
such, we add an addition step to the tool for migrating implemented
specs to remove these templates. This should automate the process in the
future.
Change-Id: Iaae10e180845b5073e1e13a4df1050b8b98d03d7
This moves the specs that were completed in Ocata
from the 'approved' directory to the 'implemented'
directory and updates the redirects file.
This also fixes a bug in the move-implemented-specs
script such that we filter out completed blueprints
which are not implemented, which was the case with
blueprint cold-migration-with-target-ocata which was
superseded.
Change-Id: I476e2e2d4bb089d901b887ce7728fc29c35a0957
This renames a couple of specs to match the actual
blueprint name in launchpad and adds redirects for the
old names.
Change-Id: I2ebfa6dc3dd9e8c8d44036f1b473397f211f83a9
The resource-providers-scheduler-db-filters spec previously contained wording
that intimated that the scheduler would begin to use the placement API for
scheduling decisions as soon as some nova-compute workers were upgraded and
began reporting inventory and allocation records to the placement API. This is
not how Ocata was released. Instead, the placement API is only queried from the
scheduler once all nova-compute workers have been upgraded to Ocata. This patch
simply records that new reality.
Change-Id: I9b9e3855b8479dd22d714149889793224909832f
In the context of virtual device role tagging at instance boot time, a
bug has caused the tag attribute to no longer be accepted starting
with version 2.33 for block_device_mapping_v2 and starting with
version 2.37 for networks. In other words, block devices can only be
tagged in 2.32 and network interfaces between 2.32 and 2.36
inclusively. This spec introduces a new API microversion that re-adds
the tag attribute to both block_device_mapping_v2 and networks.
APIImpact
Change-Id: I37cec158c6f998143842ad44f739fe2031d75eef
Implements: blueprint fix-tag-attribute-disappearing
This spec proposes that the os-aggregates REST API returns the
aggregate UUID in a new microversion so that the aggregate UUID
can be used to associate an aggregate with resource providers
in the Placement service.
APIImpact
Spec for blueprint return-uuid-from-os-aggregates-api
Change-Id: I230b329a559c41ac6e9a181b9b8c7f25a6dd1122
Currently the filter and sort query parameters can be used for
`Server List` API are not explicit defined (the filters that
can be used for non-admin users are defined).
This spec aims to propose to add whitelists for this API to
strict the filter and sort query parameters that can be used
when calling `GET /servers` API.
APIImpact
Co-Authored-By: Alex Xu <hejie.xu@intel.com>
Implements: blueprint add-whitelist-for-server-list-filter-sort-parameters
Change-Id: I7d0735b4697b22f74cf7700bca65eb38161744e8
The code in I99e7bcfe27938e5e4d50ac3005690ac1255d4c5e actually
returns the updated resource class from the PUT operation and
the response code is 200, not 204 with an empty body.
Change-Id: Ieb7a7f72f9c43886028fdaeac7a790baaea74716
This adds more detail to the spec.
Related to blueprint cells-count-resources-to-check-quota-in-api
Change-Id: I2f701b3d182b331e8533c9d10073d95f1a019ccd
Adds the concept of a tree of resource providers arranged in a
hierarchical fashion for handling resource tracking scenarios including
NUMA nodes and SRIOV physical function VF inventories.
Change-Id: I95c8bed51886bd434a025da246287c9521e520b2
blueprint: nested-resource-providers
We extend the resource-providers series to modify the
scheduler to filter compute nodes by calling the placement API
instead of retrieving all compute nodes in the deployment and
filtering on the Python side.
Change-Id: I70a48625baf36e9ac4c382bbd87baa3170198983
blueprint: resource-providers-scheduler-db-filters
As SR-IOV physical functions can not be assigned with a VLAN tag,
allow it's vlan tag to be exposed through a device metadata to the guests.
Implements blueprint sriov-pf-passthrough-neutron-port-vlan
Co-Authored-By: Artom Lifshitz <alifshit@redhat.com>
Change-Id: I9d0f4dcb0f52582704555724f4a29684896f4178
Resources should be counted to check quota in the API to minimize
coupling between the API cell and compute cells.
Related to blueprint cells-count-resources-to-check-quota-in-api
Change-Id: I6d7cb528808bb5115a8c7be4f58094bb42fb5e7f
Change how we talk to neutron to allow us in the future to implement
network aware scheduling.
This is just the continuation of the work started during newton.
This is a priority, because this blocks routed networks, which is
blocking the removal of cells v1.
blueprint prep-for-network-aware-scheduling-ocata
Change-Id: Id28509c7fae784ecc6359556a6da1bc39f2b68f7
The Nova schedular determines CPU resource utilization and instance
CPU placement based on the number of vCPUs in the flavour. A number
of hypervisors have work that is performed in the host OS on behalf
of a guest instance, which does not take place in association with
a vCPU. This is currently unaccounted for in Nova scheduling and
cannot have any placement policy controls applied.
Blueprint: libvirt-emulator-threads-policy
Change-Id: I069cd14ea89045136ae8a29c2d2f6c8c17157533
Spec to add optional parameters 'limit' and 'marker' to the
os-simple-tenant-usage endpoints for pagination.
/os-simple-tenant-usage?limit={limit}&marker={instance_uuid}
/os-simple-tenant-usage/{tenant_id}?limit={limit}&marker={instance_uuid}
APIImpact
blueprint: paginate-simple-tenant-usage
Change-Id: Ie9dc4f1c62180175433786fcf37d758762a457f0
Given that we want to get a list of ResourceProviders that can support
a request, we need to propose a REST API call having a
request body that would only return those which support that request.
Implements: blueprint resource-providers-get-by-request
Change-Id: Ibd140287e8c96ed4fa22dd9b36c274530770e207
Some operations in Nova could take a long time to complete. During this
time user token associated with this request could expire. When Nova tries
to communicate with other services using the same user token, Keystone fails
to validate the request due to expired token.
Service token will be passed along with user token to communicate with
services when dealing with long running tasks like live migration.
Keystone middlewere trusts that the service got the user token when
it was valid, don't check the expiry date of cert.
Change-Id: I7f5807ed6b998a0d88169dbbf057e22fe2cd9167
The target host can be specified when cold migrating a server.
And a new request body field called ``force`` (defaulted to False)
is added in order not to call the scheduler when providing a
destination. Setting it to 'True' makes the cold migrate operation
bypass the scheduler.
APIImpact
Change-Id: I57f3cb045a5f050931dc6c0fe9786ddd7dd0a2d7
blueprint cold-migration-with-target-ocata
xenapi is currenly involved in several OpenStack projects(e.g. nova,
neutron, ceilometer) and may be involved in other projects in the future.
By creating a new common project to hold the common xenapi modules, it will
help to reduce the code duplication between these projects; and also making
it easier to maintain, review and propose new changes to current and future
projects.
Change-Id: Ib9b06465ff2162d1f56196b1b2f1ce7a5bd45cfc
Blueprint: add-os-xenapi-library
Ian Wienand