-
Notifications
You must be signed in to change notification settings - Fork 23
Yoga zed merge #607
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Yoga zed merge #607
Conversation
eedfc4f
to
2d8b504
Compare
Weird pep8 reno failure |
Spent a while looking at that last week. I think the issue is the git tagging is all off.
But that would need your magical admin powers. I tried a few different merge/rebase style solutions but nothing quite kept all the changes properly. There might be a way but my git wrangling skills don't quite stretch that far |
8c82a9f
to
4fda999
Compare
Designate along with Redis have both been enabled to provide the multinode environment with DNSaaS. Designate has been configured to permit forwarding of queries to either GoogleDNS or CloudflareDNS.
Co-authored-by: Mark Goddard <[email protected]>
This change adds a template secrets.yml file to provide a definition for the barbican role IDs required for backend and internal TLS.
Use Tenks to provide virtual bare metal, with Kayobe helper scripts to deploy.
Needs installing in the Seed Node virtualenv.
Co-Authored-By: Alex Welsh <[email protected]>
Some dependencies for generating backend API certificates were missing but were undetected because they were fulfilled by other vault playbooks run earlier during the deployment process. This was detected when playbook hosts were extended to include separate network hosts running some OpenStack services.
* Add Ubuntu Jammy repository config * Updated docs for jammy config * Docker and UCA repository corrections * Consolidating repository config in pulp.yml * Release note for changes
Running Pulp with a number of workers based on CPUs can cause errors such as PostgreSQL connection limits on hosts with many cores: psycopg.OperationalError: connection failed: FATAL: remaining connection slots are reserved for non-replication superuser connections Cap workers to 32 until large scale testing can identify a more suitable limit.
The default Python version has changed from 3.7 to 3.11 when updating Read the Docs configuration. See the announcement of configuration file version 2 for more details [1]. [1] https://blog.readthedocs.com/migrate-configuration-v2/
Read the Docs does not perform a full clone on checkout job to reduce network data and speed up the build process. Because of this, extensions that depend on the full Git history will fail [1]. This breaks build of release notes, since reno needs Git history access. [1] https://docs.readthedocs.io/en/stable/build-customization.html#unshallow-git-clone
The Git shallow clone used on Read the Docs is configured to only fetch a specific Git branch: [remote "origin"] url = https://github.com/stackhpc/stackhpc-kayobe-config.git fetch = +refs/heads/stackhpc/yoga:refs/remotes/origin/stackhpc/yoga This prevents subsequent `git fetch` commands from fetching all remote branches. Use `git remote set-branches origin '*'` to revert to the default Git fetch configuration.
The wildcard glob matched files in the current directory which resulted in `git fetch` failing with: fatal: couldn't find remote ref refs/heads/LICENSE
Use an explicit list of remote branches with release notes instead of a wildcard which is behaving differently on Read the Docs than locally. Also remove true fallback to fail early.
Uses the same os_release template as the ci-multinode environment.
It prevents NetworkManager from automatically running DHCP on unconfigured ethernet devices and allows connections with static IP addresses to be brought up even on ethernet devices with no carrier.
Bumped focal repos to fix unmet dependencies on container build. Built new Magnum containers including fix for missing Octavia load balancer health monitors.
Fixes bug 2015393 - missing Octavia health monitors on cluster spawn.
This requires stackhpc.cephadm 1.14.0.
56f7c47
to
ec3063b
Compare
ec3063b
to
ed8adaa
Compare
Image builds were previously failing due to the runners being provisioned with CS8 (Zed requires python > 3.6 which CS8 does not support). This change edits the workflow to use new RL9 images, which can build Zed container images. The workflow also had some CS8-specific customisations which have been removed
f5ccd6f
to
cc2cf4d
Compare
Closing in favour of #620 due to my mistake rebasing this branch on stackhpc/zed |
Syncing latest stackhpc/yoga into Zed. Only real conflicts were in pulp.yml where formatting was changed to reduce repetition.