Permalink
Checking mergeability…
Don’t worry, you can still create the pull request.
Comparing changes
Choose two branches to see what’s changed or to start a new pull request.
If you need to, you can also or
learn more about diff comparisons.
Open a pull request
Create a new pull request by comparing changes across two branches. If you need to, you can also .
Learn more about diff comparisons here.
...
- 17 commits
- 11 files changed
- 7 contributors
Commits on Jun 8, 2026
-
* Update neutron from branch 'master' to 1fcf08a48eb5e19c8127b87053f6ecde99136145 - Merge "Remove ssl options" - Remove ssl options These options from oslo_service were used by the api workers, which were removed when neutron-server was removed. Change-Id: Id07194b432401961137c7635a2222dad26182499 Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com> -
* Update glance from branch 'master' to b7919f05c50c7e5120530458ab7eda939a29ab0f - Merge "Fix vague 415 responses for unsupported Content-Type headers" - Fix vague 415 responses for unsupported Content-Type headers Glance returned HTTP 415 when clients sent an unsupported Content-Type, but the response body did not explain the real problem. Pass the error message via WebOb's detail= field for image upload, stage, and PATCH so users see the actual reason (for example, invalid upload type or allowed PATCH media types). Assisted-by: Cursor (claude-4.5-sonnet) Closes-Bug: #2152232 Change-Id: I49d786433ef21ac7407edb2865fe0708a6d5a309 Signed-off-by: PranaliD <pdeore@redhat.com> -
* Update glance from branch 'master' to 0cd329e8898fbf59502befe52f5b8800ab43cf97 - Merge "Rally: use raw disk_format for synthetic fake.img" - Rally: use raw disk_format for synthetic fake.img Glance inspects the upload stream; ~/.rally/extra/fake.img is not qcow2, so claiming qcow2 triggers InvalidImageData and HTTP 415 on GlanceImages.create_and_list_image. Use disk_format raw for that subtask only; keep qcow2 for the CirrOS HTTP workload. Assisted-By: Cursor (claude-4.5-sonnet) Change-Id: I743131df0d8c518ec00cf5958136cd64d76afb30 Signed-off-by: PranaliD <pdeore@redhat.com> -
* Update nova from branch 'master' to 9faf778a59154c2caa318cbdff19b183dd0f5d84 - Merge "Use common utility to parse Host header" - Use common utility to parse Host header Change-Id: Icb21762676a810ed2141be34ab8d886198dab322 Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com> -
* Update openstack-ansible-os_designate from branch 'master' to de02bd9e3d0362d5912a919e11e8ebbdf98a5b8d - Start using uWSGI role Move service to use uWSGI role instead of iternal task for uwsgi deployment. This aims to ease the maintenance of uWSGI and speedup metal deployments as the same uwsgi environment will be used across all services. Change-Id: I1e7d6a9fa87f18ce2515e28a5f3df4f6ad01b485 Signed-off-by: Ivan Anfimov <lazekteam@gmail.com> -
* Update openstack-ansible-os_designate from branch 'master' to 46af54fc3b170ce6873f836de343b3c36e91153d - Merge "Do not remove policy.yaml file" - Do not remove policy.yaml file oslo.policy can not handle policy file removal. As a result, if policy overrides were defined at some point, but then removed, causes service outage. While we could add a handler trigger to restart the service on policy removal, it's better to simplify the logic and always place an empty policy.yaml even if no overrides are defined. Change-Id: I0ad1d4c14eec381569e38af36065c6d3b6955f53 Co-authored-by: Dmitriy Rabotyagov <dmitriy.rabotyagov@cleura.com> Signed-off-by: Ivan Anfimov <lazekteam@gmail.com> -
* Update nova from branch 'master' to 6bbe904c611266d9b588e129aea4dfc24bba949a - Merge "Use oslo.utils to parse cell URL" - Use oslo.utils to parse cell URL ... to reduce duplicate logic. Change-Id: I6b24e3841a3b84e974391134ccfbabcf6fabb137 Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com> -
* Update ironic from branch 'master' to a58fe7d548e3b874b7538d6272b2a73f3a0d4b94 - Fix kernel parameter parsing for quoted values and whitespace The kernel parameter validator introduced in commit c6c91d649 has three bugs in the KernelParameterTransformer class: 1. The quoted_value() method tries to access items[0].value but items[0] is already a plain string (returned by value_with_spaces transformer), not a lark Token. This causes all quoted parameter values (e.g. sshkey="ssh-rsa ... user@host") to fail with "'str' object has no attribute 'value'". 2. KernelParameter.__str__() uses self.value.value directly, bypassing ParameterValue.__str__() which wraps space-containing values in quotes. This breaks roundtrip rendering of quoted values. 3. parse() does not strip surrounding whitespace from the input, causing failures when config values have trailing spaces or newlines from template expansion or oslo.config. Assisted-By: Claude Opus 4.6 Change-Id: I9a398d557c3f87c070ffa10e317f0df613b98b6b Signed-off-by: Riccardo Pittau <elfosardo@gmail.com> -
* Update requirements from branch 'master' to df8c1d8697a531140de152d548ae02b731e1a235 - update constraint for os-traits to new release 3.7.0 meta: version: 3.7.0 meta: team: nova meta: diff-start: - meta: series: independent meta: branch: master meta: release-type: release meta: pypi: yes meta: first: no meta: release:Author: Balazs Gibizer <gibi@redhat.com> meta: release:Commit: Balazs Gibizer <gibi@redhat.com> meta: release:Change-Id: I2291b42c9e31244e51b6ddffa1cbff59fa66e174 meta: release:Code-Review+1: ribaudr <rene.ribaud@gmail.com> meta: release:Code-Review+2: Elod Illes <elod.illes@est.tech> meta: release:Workflow+1: Elod Illes <elod.illes@est.tech> Change-Id: I4642ba4acc0150df75fdfdfad8129d8e00d83db1 Signed-off-by: OpenStack Proposal Bot <openstack-infra@lists.openstack.org> Generated-By: openstack/project-config:roles/copy-release-tools-scripts/files/release-tools/update_constraints.sh -
* Update neutron from branch 'master' to fae346375e8e72a133e8717fdf2a554cbec069a0 - Merge "Add compute1 node to BGP multinode tempest job" - Add compute1 node to BGP multinode tempest job Add a second compute node (compute1) to the neutron-ovn-bgp-tempest-multinode job so that VMs can be scheduled on a node that does not run the OpenStack control plane. Compute1 is dual-homed to both leaf switches, runs FRR with BFD, and the OVN agent with the ovn-bgp extension, matching the controller's BGP setup. Changes: - Nodeset: add compute1 to nodes, compute, subnode, and new zbgp_compute group; rename to neutron-five-node-noble-bgp. The zbgp_compute group name is chosen to sort after "subnode" alphabetically so its group-vars override the subnode group-vars from the devstack base job. - Job: move spine config from subnode to tempest group-vars; add zbgp_compute group-vars for compute1. Both tempest and zbgp_compute group-vars include the base subnode devstack_localrc settings (passwords, host addresses) since their alphabetical precedence replaces the subnode group-vars entirely. - Job: add zuul_copy_output for FRR, OVN, and OVS logs. - Playbook: merge controller and compute1 tunnel plays into a single play targeting the compute group with a per-host config map for VXLAN key/VLAN prefixes. - Playbook: fix FRR start guard to use a last_compute_node variable (last compute subnode) instead of relying on Ansible group ordering. Generate SSH keys on all compute nodes so the last one can start FRR on leaf/spine. - Leaf FRR: add compute1-port as a downlink neighbor. - Docs: update topology diagram, node descriptions, and VXLAN wiring table. - Test test_reuse_ip_address_with_other_fip_on_other_router is skipped because of a core ovn issue. Depends-On: https://review.opendev.org/c/openstack/neutron/+/991433 Assisted-By: Claude Opus 4.6 Change-Id: I5f981f1838c3b51b6a7d2c9032192fc6c5c88f15 Signed-off-by: Eduardo Olivares <eolivare@redhat.com> -
* Update openstack-helm from branch 'master' to 11490b020c3f9ba6fe2fad73ab55d85518710efb - libvirt: Use emptyDir for generated /etc/ceph The libvirt daemonset stores generated Ceph configuration and keyrings in the etcceph volume. That volume is populated by the Ceph placement init containers and consumed by the libvirt container in the same pod, so it does not need host persistence. Use an emptyDir for etcceph instead of a hostPath under /var/lib/openstack-helm/compute/libvirt. This keeps the generated Ceph files inside the pod mount namespace and avoids stale Ceph bind mounts accumulating on the host during libvirt pod restarts. Change-Id: Iccf55ffaf2c82ccb1e9367bc554af107f054cf7c Signed-off-by: Dong Ma <winterma.dong@gmail.com> -
* Update kolla from branch 'master' to 7477b2244ee525ffd6effeb4229608991c639fc8 - keepalived: Install mariadb client package Needed-By: https://review.opendev.org/c/openstack/kolla-ansible/+/992020 Related-Bug: #2155865 Change-Id: I1c67f6b0f4780b7caf83ccf95a20861d9fefe20b Signed-off-by: Michal Nasiadka <mnasiadka@gmail.com> -
* Update telemetry-tempest-plugin from branch 'master' to aaa84d78f1480a499211096ed78454480c4bb923 - Merge "Skip heat stack cleanup when the test fails early" - Skip heat stack cleanup when the test fails early In case a test case fails quite early (during setUp) the environment variable might not be present. Change-Id: Id16552daef74f418b20fa1fc08c5795d0f593adc Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com> -
* Update charm-openstack-dashboard from branch 'master' to 6fb0b742753a642e3e660d471cbc28d9de644642 - Revert the default upload limit to Jammy's limit There was a change in Apache2 upstream and Ubuntu packaging about the default value for LimitRequestBody between jammy and noble. On noble based deployment, any >1 GiB image upload doesn't succeed and Apache2 records 413 Content Too Large out of the box. We revert the value for LimitRequestBody to jammy's one to keep the previous behavior and be backward compatible. Some alternatives were considered: - Switch to HORIZON_IMAGES_UPLOAD_MODE=direct instead of accepting the upload at Horizon + This changes the traffic flow from a web browser, and Glance API endpoint must be reachable by the browser + It may not be backward compatible with existing deployment including a firewall setup between a web browser and the OpenStack environment + A browser can block the upload as CORS (Cross-Origin Resource Sharing) since Horizon and Glance have different vips in Charmed OpenStack and FQDN - Have a "sensitive" limit for LimitRequestBody instead of going back to unlimited + It's impossible to come up with the sensible value in Horizon. Glance charm has the "image-size-cap" option, but the dashboard charm cannot see the value set for the option in the current structure of relations in Charmed Openstack. + Even if the dashboard charm has access to the value, the default there is 1TB, which is gigantic already. I don't see much difference between 1TB and unlimited to a reasonable size of units for Horizon units. + We could create a new config option in the dashboard charm as "image-size-cap" to set LimitRequestBody by the admin. But it may be an overkill and lead confusion that the image size limit needs to be set in both Glance and Horizon separately. Closes-Bug: #2155246 Signed-off-by: Nobuto Murata <nobuto.murata@canonical.com> Change-Id: If8ceffad7afa4b2ad31b0bf83c2b9ac33ef141fe -
* Update openstack-helm from branch 'master' to cc1f18398e6de6e93a5dd2bda92be97bbb2d4c3c - Merge "[cinder] Add optional StatefulSet for cinder-volume" - [cinder] Add optional StatefulSet for cinder-volume The static host value (conf.cinder.DEFAULT.host) keeps the cinder-volume service identity stable across pod restarts, but it is only safe with a single replica: every pod registers under the same host name and competes for the same RPC queue, periodic tasks, and (without a Tooz coordination backend plus [DEFAULT] cluster) uncoordinated locks. Add an optional StatefulSet manifest for cinder-volume, gated by manifests.statefulset_volume (disabled by default). When enabled, the existing deployment_volume manifest is suppressed and each pod gets a stable, unique hostname (cinder-volume-0, cinder-volume-1, ...). The cinder-volume.sh script now derives the host from ${HOSTNAME} and appends a per-pod --config-file when conf.cinder.DEFAULT.host is left empty, so each replica registers under its own pod name. Comments are added in values.yaml next to the host field and the new manifests.statefulset_volume flag to document the single-replica caveat and how to switch to the StatefulSet for predictable hostnames. The deploy-charts cinder role enables manifests.statefulset_volume and nulls conf.cinder.DEFAULT.host so the cinder CI job exercises the new StatefulSet path. This depends on the helm-toolkit change in the preceding manila commit that omits oslo.config keys whose value is nil (without it, the nil host would render as the literal string "<no value>" and oslo.config would reject it). Signed-off-by: Vladimir Kozhukalov <kozhukalov@gmail.com> Change-Id: I94ced1bdbdf4f55f7f169e57e9f6393407e1a989 -
* Update neutron from branch 'master' to 1bce68e8b03ccf09d1bab83efc7ca6a4f65496f9 - Merge "Validate [DEFAULT] setproctitle while loading config files" - Validate [DEFAULT] setproctitle while loading config files ... to detect unsupported values early. Also use the native interface to document available choices. Change-Id: I43848aebddc1819101f969a452560b5b37dc839b Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com> -
* Update placement from branch 'master' to 3cba66c70ce49e31e582c8d03e945017f38fcf11 - Merge "Update python versions" - Update python versions Python 3.10 was removed from the tested runtimes during this cycle. Also add python 3.14 which will be required later. Change-Id: I9d61407180c8db04778bd3d3dd58df965a353257 Signed-off-by: Takashi Kajinami <kajinamit@oss.nttdata.com>
Loading
This comparison is taking too long to generate.
Unfortunately it looks like we can’t render this comparison for you right now. It might be too big, or there might be something weird with your repository.
You can try running this command locally to see the comparison on your machine:
git diff master...master