Repository navigation
zuul: move the stable lanes to 11.0.2 - #2978
Merged
Merged
Conversation
Make OSISM 11.0.2 the stable release the testbed lanes deploy. - The stable deploy lanes move from 11.0.1 to 11.0.2. - The major upgrade lane goes from 10.3.0 to 11.0.2. - The minor update lane goes from 11.0.1 to 11.0.2 instead of from 10.2.0 to 10.3.0. 11.0.2 is the first release with an in-series predecessor. Starting from 11.x, it needs ceph_version: tentacle like the deploy lanes: deploy.yml defaults to reef. - testbed-update-stable-current and testbed-upgrade-stable-next start from 11.0.2. Their latest targets are unchanged. Assisted-by: Claude:claude-opus-5-5 Signed-off-by: Roger Luethi <luethi@osism.tech>
upgrade-manager.sh decides whether upgrade-services.sh moves Ceph by writing SKIP_CEPH_UPGRADE into /opt/manager-vars.sh, but only when the target manager version is "latest", where ceph_version_next selects it. For a stable target the variable stays unset, because the release brings its own Ceph pin, and upgrade-services.sh defaults it to false. On a ceph-ansible testbed that runs the ceph-ansible upgrade, as it always has. On a cephadm testbed it trips the guard that refuses the ceph-ansible upgrade there: moving Ceph under cephadm takes "ceph orch upgrade", which OSISM does not provide yet. No stable lane started on cephadm until testbed-update-stable moved to 11.0.1 -> 11.0.2. deploy.yml picks cephadm for any manager version from 11.0.0 on, so that lane now deploys cephadm and fails at the guard before any service is upgraded. For a stable target on cephadm, write SKIP_CEPH_UPGRADE=true, so the manager, infrastructure, OpenStack and monitoring are upgraded and Ceph stays as deployed. CEPH_STACK is read from /opt/manager-vars.sh in a subshell, so the variables it sets do not override the ones this script takes from its arguments. The guard in upgrade-services.sh stays in place for anyone who asks for a Ceph move explicitly. This loses no coverage today: 11.0.1 and 11.0.2 pin the same Ceph image, 20.2.4-r1. Once an 11.x release moves that pin, the lane will leave the cluster on the old image until there is a cephadm upgrade path to exercise. The 10.x -> 11.x upgrade lane deploys ceph-ansible and keeps it, so it is unaffected. Assisted-by: Claude:claude-opus-5-5 Signed-off-by: Roger Luethi <luethi@osism.tech>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Make OSISM 11.0.2 the stable release the testbed lanes deploy.
ceph_version: tentaclelike the deploy lanes:deploy.ymldefaults to reef.testbed-update-stable-currentandtestbed-upgrade-stable-nextstart from 11.0.2. Theirlatesttargets are unchanged.Ceph in the minor update lane
The update lane is the first stable lane that starts on cephadm (
deploy.ymlpicks cephadm from 11.0.0 on). With only the lane change, it failed before upgrading anything:upgrade-manager.shsetSKIP_CEPH_UPGRADEonly for alatesttarget, so for a stable targetupgrade-services.shdefaulted it to false and refused the ceph-ansible upgrade on a cephadm cluster. Moving Ceph under cephadm takesceph orch upgrade, which OSISM does not provide yet.The second commit writes
SKIP_CEPH_UPGRADE=truefor a stable target on cephadm. The lane upgrades the manager, infrastructure, OpenStack and monitoring, and leaves Ceph as deployed. That costs no coverage today: 11.0.1 and 11.0.2 pin the same Ceph image, 20.2.4-r1. Once an 11.x release moves that pin, the lane will keep the old image until there is a cephadm upgrade path to test.Other lanes are unaffected:
latesttarget already setceph_version_next: skip.The lanes need 11.0.2 in osism/release and in the seed image first:
🤖 Generated with Claude Code