-
Notifications
You must be signed in to change notification settings - Fork 112
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
Override Ceph version #2574
Override Ceph version #2574
Conversation
Thanks for the PR! ❤️ |
de0ac66
to
9d1f1cf
Compare
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/0519f4329a914189ae7f5384c6390013 ✔️ openstack-k8s-operators-content-provider SUCCESS in 2h 04m 20s |
d437a63
to
a83a33e
Compare
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/51e82232d4f34df5a6eb74ec7a7e5c14 ✔️ openstack-k8s-operators-content-provider SUCCESS in 1h 30m 34s |
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/68ae94ace71b4143bd4c78b6529f961d ✔️ openstack-k8s-operators-content-provider SUCCESS in 1h 30m 42s |
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/4278fce779f24a239cc8073ca3980e52 ✔️ openstack-k8s-operators-content-provider SUCCESS in 1h 43m 15s |
c1777aa
to
8427cbc
Compare
@@ -55,7 +55,9 @@ | |||
{% if not cifmw_cephadm_default_container %}--image {{ cifmw_cephadm_container_ns + '/' + cifmw_cephadm_container_image + ':' + cifmw_cephadm_container_tag|string }} \{% endif %} | |||
bootstrap \ | |||
--skip-firewalld \ | |||
{% if not cifmw_cephadm_prepare_host %} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@fultonj looks like from Squid+ we don't need to --skip-prepare-host
, as we already have everything in place when we reach this point (network, lvdevices, podman) [1].
I would try a testproject with this change to make sure we don't hit any side effect.
[1] https://github.com/ceph/ceph/blob/squid-release/src/cephadm/cephadm.py#L4494
- cifmw_cephadm_repository_override | bool | ||
become: true | ||
ansible.builtin.dnf: | ||
name: centos-release-ceph-{{ cifmw_cephadm_version }} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
dlrn embeds the Centos-Storage-Sig repos, and by installing this package we can simply override the Ceph related sections, bumping the version to a newer release.
zuul.d/edpm_multinode.yaml
Outdated
@@ -343,6 +343,15 @@ | |||
dashboard_enabled: true | |||
cephfs_enabled: true | |||
ceph_nfs_enabled: true | |||
# Override the Ceph container image and deploy Squid | |||
cifmw_cephadm_container_ns: "quay.io/ceph" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@fultonj @tosky I was thinking to switch this multinode-hci-edpm
to Ceph Squid
, so we have a way to do early testing upstream.
- Does it make sense to do this here? It would help to perform a deployment sanity check, like we are going to do in devstack (but in that case I'm not sure we're going to backport squid to Antelope)
- If you're positive about Squid, do we want to add here a stable tag (e.g.
v19.2
) to avoid testing thelatest
build?
I can simply remove this part here and only target something downstream. I wanted to get your opinion and feedback first because I'm potentially going to do the same thing in install_yamls
[1] and target this new Ceph versions for tempest jobs w/ Glance, Cinder and Manila.
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/a7e7229850874053b83bcd2f133c10c7 ✔️ openstack-k8s-operators-content-provider SUCCESS in 1h 56m 02s |
recheck |
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/4fa7d899abfb4875a33621fcd72cd591 ✔️ openstack-k8s-operators-content-provider SUCCESS in 2h 20m 35s |
The failure in RDO is not related to this patch.
https://softwarefactory-project.io/zuul/t/rdoproject.org/build/c4f534a1c131475082e25256078ef1d6 RDO is green otherwise, in particular Thus, I'm +2 to merge this. Though the unrelated NNCP issue seems to be blocking us. |
recheck |
we should setup a testproject to make sure we don't hit any side effect. We shouldn't but we can take some time to do some extra testing. |
@fultonj from internal testing looks good and we can land this patch if you're ok with it. |
zuul.d/edpm_multinode.yaml
Outdated
@@ -343,6 +343,15 @@ | |||
dashboard_enabled: true | |||
cephfs_enabled: true | |||
ceph_nfs_enabled: true | |||
# Override the Ceph container image and deploy Squid |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do we really need the values here that are matching the defaults? Not really agains it but this means that we need to touch a few places to bump the version.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
mmm I wanted a place where we can have a well defined list of parameters that allow us to test the next version.
For this reason I can revert back the defaults under defaults/main.yaml
to match Reef
, and leave here, at job level, the overrides to install Squid
.
How does it sounds?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done in the meantime, so you can check what I had in mind.
We used to support N+1 server version with external Ceph, and this policy is still adopted. For this reason we need a way to do early testing in CI. This patch adds the ability to override the CentOS Ceph repository and install the target cephadm release passed as input. Signed-off-by: Francesco Pantano <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/approve
/lgtm
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: fultonj The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/4f10fa0c37784564978af0478a85cd5e ✔️ openstack-k8s-operators-content-provider SUCCESS in 2h 11m 45s |
recheck |
Build failed (check pipeline). Post https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/0941d1757fa148afbe2856c86dcb075f ✔️ openstack-k8s-operators-content-provider SUCCESS in 2h 18m 26s |
recheck |
49034ad
into
openstack-k8s-operators:main
/cherry-pick 18.0-fr1 |
@fmount: new pull request created: #2641 In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
We used to support
N+1
server version with externalCeph
, and this policy is still adopted.For this reason we need a way to do early testing in CI.
This patch adds the ability to override the
CentOS
Ceph
repository andinstall the target
cephadm
release passed as input.Jira: https://issues.redhat.com/browse/OSPRH-10666