Use MultiTemplateDir for the support of multi-segment dir in template by amartyasinha · Pull Request #1096 · openstack-k8s-operators/nova-operator

amartyasinha · 2026-04-09T13:44:23Z

With the restructure of the templates dir to have multi-segment dir, this commit ensures the proper templating from these dirs.

lib-common change introducing MultiTemplateDir: openstack-k8s-operators/lib-common#682

openshift-ci · 2026-04-09T13:44:29Z

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

softwarefactory-project-zuul · 2026-04-09T13:54:50Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/f92eab4ff0494c3c974f4b47862cc437

❌ openstack-meta-content-provider FAILURE in 9m 22s
⚠️ nova-operator-kuttl SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode-ceph SKIPPED Skipped due to failed job openstack-meta-content-provider

softwarefactory-project-zuul · 2026-04-10T11:40:49Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/bd483d15a9304711826243d8e3546a72

❌ openstack-meta-content-provider FAILURE in 9m 49s
⚠️ nova-operator-kuttl SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode-ceph SKIPPED Skipped due to failed job openstack-meta-content-provider

bogdando · 2026-04-10T12:55:11Z

So for Kinds that already had explicit cases in #1090, GetTemplateSubpath is intended to reproduce the same nova/... paths. For Kinds that previously fell through the old default strings.ToLower(kind), this PR implicitly switches them to the GetTemplateSubpath rule (NovaCell -> nova/cell instead of novacell). The
PR, as shown, does not:

document that audit,
add per-Kind overrides, or
move template trees for those Kinds (create templates/nova/cell/... and retire templates/novacell/... etc)

amartyasinha · 2026-04-13T06:29:02Z

This PR intends to remove the InstanceType overloading (done in #1090), and leave InstanceType to only have kind as value. MultiTemplateDir is an optional parameter, when used will take over to set the subdir during templating. Only when this parameter is set, GetTemplateSubpath will be executed.

Regarding

document that audit,

I did not get where to audit this change. Should I amend the commit message to include all details, or add them somewhere else?

add per-Kind overrides, or

Sorry, did not get it. Could you please explain it a bit (maybe an example would be helpful)

move template trees for those Kinds (create templates/nova/cell/... and retire templates/novacell/... etc)

IIUC, #1090 has already moved template trees for all kinds as templates/nova/<kind>/.., and we do not have templates/nova<kind>/... Please correct me if I've missed something.

gibizer

You can try to run a build and test in CI by temporarily adding the new lib-common hash to the go.mod file so the build can see the new field.

gibizer · 2026-04-16T07:11:09Z

 			Type:               util.TemplateTypeConfig,
-			InstanceType:       getTemplateInstanceType(instance),
+			InstanceType:       gvk.Kind,
+			MultiTemplateDir:   util.GetTemplateSubpath(gvk.Group, gvk.Kind),


I think this is where the change go too far by trying to generate the path instead of just defining the path per service. Trying to support the generation is just unnecessary complexity. Simply define the path per service in the nova-operator then we can avoid this complexity.

ack! I thought we need to generate path from kind, i.e. NovaAPI -> nova + api. Since we can define the path directly instead of generating, I'll remove the generation functions from lib-common, and close the CRD spec modification PR.

defined the path instead of generating them

gibizer · 2026-04-16T07:12:35Z

+			Namespace:        instance.Namespace,
+			Type:             util.TemplateTypeScripts,
+			InstanceType:     instance.GetObjectKind().GroupVersionKind().Kind,
+			MultiTemplateDir: "nova/nova-manage",


Yepp this is the correct use of the new interface.

Explicit is better than implicit.
Simple is better than complex.

gibizer · 2026-04-16T07:19:54Z

 	templateParameters map[string]any,
 	extraData map[string]string, cmLabels map[string]string,
 	additionalTemplates map[string]string,
 	withScripts bool,


You probably need a new param here to pass the template dir.

Or you can try to change the instance client.Object to be a specific interface type where the interface defines a getter for the template dir and then implement that
interface function for our Objects

Or you can add the getter for the ReconcileBase and then implement it in each Reconciler to return the service specific path.

added a new param templateDir.

openshift-ci · 2026-04-16T08:25:06Z

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: amartyasinha
Once this PR has been reviewed and has the lgtm label, please ask for approval from gibizer. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

softwarefactory-project-zuul · 2026-04-16T08:40:06Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/543c38b98e54484585ca5047ccff327d

❌ openstack-meta-content-provider FAILURE in 9m 11s
⚠️ nova-operator-kuttl SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode-ceph SKIPPED Skipped due to failed job openstack-meta-content-provider

softwarefactory-project-zuul · 2026-04-16T12:36:17Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/de67ec79fde947fdb7710c52538dff49

✔️ openstack-meta-content-provider SUCCESS in 3h 47m 10s
❌ nova-operator-kuttl FAILURE in 47m 24s
✔️ nova-operator-tempest-multinode SUCCESS in 2h 07m 44s
✔️ nova-operator-tempest-multinode-ceph SUCCESS in 3h 07m 09s

amartyasinha · 2026-04-16T13:08:44Z

recheck nova-operator-kuttl

gibizer

Overall looks good to me.

gibizer · 2026-04-16T14:15:57Z

-		InstanceType: getTemplateInstanceType(instance),
+		InstanceType: instance.GetObjectKind().GroupVersionKind().Kind,
 		Labels:       labels,
 		CustomData:   data,


I guess we are not loading any template from file here hence we don't need to pass a templateDir. If I'm correct then a comment would be nice here about it.

In lib-common, we do not check InstanceType (old logic) or MultiTemplateDir (new change) if template type is none.
Ref: https://github.com/amartyasinha/lib-common/blob/135bf0acdeafc623f9f9f855ca6850748664ff69/modules/common/util/template_util.go#L297

Will add a comment in there in code to clarify.

gibizer · 2026-04-16T14:18:27Z


 replace k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20250627150254-e9823e99808e //allow-merging
+
+replace github.com/openstack-k8s-operators/lib-common/modules/common => github.com/amartyasinha/lib-common/modules/common v0.0.0-20260416075014-135bf0acdeaf


yepp this way we can get CI results. After the lib-common change lands we need to remove this line but that is a small price to pay to have early CI results

softwarefactory-project-zuul · 2026-04-16T18:07:14Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/31acb1fdb20342ca9bd0d672e26cc7d0

✔️ openstack-meta-content-provider SUCCESS in 2h 57m 37s
❌ nova-operator-kuttl RETRY_LIMIT in 36m 59s
✔️ nova-operator-tempest-multinode SUCCESS in 2h 20m 35s
✔️ nova-operator-tempest-multinode-ceph SUCCESS in 2h 41m 38s

amartyasinha · 2026-04-17T04:18:39Z

recheck nova-operator-kuttl

bogdando · 2026-04-17T08:48:42Z

 			Type:               util.TemplateTypeNone,
-			InstanceType:       getTemplateInstanceType(instance),
+			InstanceType:       instance.GetObjectKind().GroupVersionKind().Kind,
+			MultiTemplateDir:   "nova/metadata",


it seems that other callers with TemplateTypeNone do NOT provide MultiTemplateDir.
Why this one is special?

Please evaluate it for consistency with lib-common behavior and other call sites

Thanks for the catch! removed MultiTemplateDir here as this specific case is of TemplateTypeNone.

bogdando · 2026-04-17T09:02:55Z

 			Namespace:          instance.GetNamespace(),
 			Type:               util.TemplateTypeConfig,
-			InstanceType:       getTemplateInstanceType(instance),
+			InstanceType:       instance.GetObjectKind().GroupVersionKind().Kind,


For a NovaCell instance, lib-common will look under a subdir derived from "novacell", not "nova/compute". Is that expected?

looks about right, sorry

bogdando · 2026-04-17T09:10:58Z

 	configName := instance.GetName() + "-compute-config"
 	err := r.GenerateConfigs(
-		ctx, h, instance, configName, &hashes, templateParameters, extraData, cmLabels, map[string]string{},
+		ctx, h, instance, configName, &hashes, templateParameters, extraData, cmLabels, map[string]string{}, "nova/compute",


ok, I see that this resolved the 'novacell' mismatching behavior I thought still takes place

amartyasinha · 2026-04-17T10:53:12Z

lib-common bump PR: #1088

softwarefactory-project-zuul · 2026-04-17T10:57:59Z

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://softwarefactory-project.io/zuul/t/rdoproject.org/buildset/5dd4abb4e8b844e8a0187f87ccbc6b30

❌ openstack-meta-content-provider FAILURE in 8m 44s
⚠️ nova-operator-kuttl SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode SKIPPED Skipped due to failed job openstack-meta-content-provider
⚠️ nova-operator-tempest-multinode-ceph SKIPPED Skipped due to failed job openstack-meta-content-provider

stuggi · 2026-04-17T11:56:56Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

amartyasinha · 2026-04-20T06:11:04Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

The bump has happened in that PR. What's the procedure to get that merged?

stuggi · 2026-04-20T06:50:52Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

The bump has happened in that PR. What's the procedure to get that merged?

the lib-common PR got merged, but the reference in the go.mod of the nova-operator was not bumped yet. you'd have to do that. https://github.com/openstack-k8s-operators/nova-operator/blob/main/go.mod#L15-L17

amartyasinha · 2026-04-20T06:56:22Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

The bump has happened in that PR. What's the procedure to get that merged?

the lib-common PR got merged, but the reference in the go.mod of the nova-operator was not bumped yet. you'd have to do that. https://github.com/openstack-k8s-operators/nova-operator/blob/main/go.mod#L15-L17

Yes, get that. This automation PR [1] is bumping the lib-common changes in nova-operator. It now points to the latest commit id 81c71b3 of lib-common. So, I was wondering will you merge that PR or @gibizer ?

[1] https://github.com/openstack-k8s-operators/nova-operator/pull/1088/changes

stuggi · 2026-04-20T07:32:26Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

The bump has happened in that PR. What's the procedure to get that merged?

the lib-common PR got merged, but the reference in the go.mod of the nova-operator was not bumped yet. you'd have to do that. https://github.com/openstack-k8s-operators/nova-operator/blob/main/go.mod#L15-L17

Yes, get that. This automation PR [1] is bumping the lib-common changes in nova-operator. It now points to the latest commit id 81c71b3 of lib-common. So, I was wondering will you merge that PR or @gibizer ?

[1] https://github.com/openstack-k8s-operators/nova-operator/pull/1088/changes

yes, when tests passed, we can approve it

amartyasinha · 2026-04-20T07:34:12Z

lib-common bump PR: #1088

you'd have to bump lib-common in go.mod and api/go.mod in this pr, or wait for a next bump to happen.

The bump has happened in that PR. What's the procedure to get that merged?

the lib-common PR got merged, but the reference in the go.mod of the nova-operator was not bumped yet. you'd have to do that. https://github.com/openstack-k8s-operators/nova-operator/blob/main/go.mod#L15-L17

Yes, get that. This automation PR [1] is bumping the lib-common changes in nova-operator. It now points to the latest commit id 81c71b3 of lib-common. So, I was wondering will you merge that PR or @gibizer ?
[1] https://github.com/openstack-k8s-operators/nova-operator/pull/1088/changes

yes, when tests passed, we can approve it

ack! thanks

With the restructure of the templates dir to have multi-segment dir, this commit ensures the proper templating from these dirs. Signed-off-by: Amartya Sinha <amsinha@redhat.com>

amartyasinha · 2026-04-21T08:22:42Z

/retest

openshift-ci Bot added the do-not-merge/work-in-progress label Apr 9, 2026

amartyasinha force-pushed the restructure-template branch from 833acea to 4205714 Compare April 10, 2026 11:29

amartyasinha marked this pull request as ready for review April 14, 2026 07:18

openshift-ci Bot removed the do-not-merge/work-in-progress label Apr 14, 2026

openshift-ci Bot requested review from mrkisaolamb and stuggi April 14, 2026 07:18

amartyasinha changed the title ~~[DNM] Use MultiTemplateDir for the support of multi-segment dir in template~~ Use MultiTemplateDir for the support of multi-segment dir in template Apr 14, 2026

amartyasinha marked this pull request as draft April 14, 2026 07:19

openshift-ci Bot added the do-not-merge/work-in-progress label Apr 14, 2026

amartyasinha changed the title ~~Use MultiTemplateDir for the support of multi-segment dir in template~~ [DNM] Use MultiTemplateDir for the support of multi-segment dir in template Apr 14, 2026

openshift-ci Bot removed the do-not-merge/work-in-progress label Apr 14, 2026

This was referenced Apr 14, 2026

api: add openStackServiceIdentity to Nova service and template specs #1099

Closed

util: Add MultiTemplateDir to support multi-segment path in template openstack-k8s-operators/lib-common#682

Merged

gibizer requested changes Apr 16, 2026

View reviewed changes

openshift-ci Bot assigned gibizer Apr 16, 2026

amartyasinha force-pushed the restructure-template branch from 4205714 to ae96eb3 Compare April 16, 2026 08:25

amartyasinha force-pushed the restructure-template branch 2 times, most recently from cb7af73 to 8e6c407 Compare April 16, 2026 08:29

gibizer reviewed Apr 16, 2026

View reviewed changes

amartyasinha force-pushed the restructure-template branch from ff25b96 to ec94f8a Compare April 16, 2026 15:08

openshift-ci Bot added the do-not-merge/work-in-progress label Apr 17, 2026

amartyasinha requested a review from gibizer April 17, 2026 06:50

bogdando reviewed Apr 17, 2026

View reviewed changes

amartyasinha force-pushed the restructure-template branch 2 times, most recently from 3f89221 to ccbea21 Compare April 17, 2026 09:37

amartyasinha marked this pull request as ready for review April 17, 2026 09:48

openshift-ci Bot removed the do-not-merge/work-in-progress label Apr 17, 2026

openshift-ci Bot requested a review from abays April 17, 2026 09:48

amartyasinha changed the title ~~[DNM] Use MultiTemplateDir for the support of multi-segment dir in template~~ Use MultiTemplateDir for the support of multi-segment dir in template Apr 17, 2026

amartyasinha force-pushed the restructure-template branch 2 times, most recently from a13603e to 809d07c Compare April 17, 2026 10:48

amartyasinha force-pushed the restructure-template branch from 809d07c to 06fa137 Compare April 21, 2026 07:42

Use MultiTemplateDir for the support of multi-segment dir in template

defa882

With the restructure of the templates dir to have multi-segment dir, this commit ensures the proper templating from these dirs. Signed-off-by: Amartya Sinha <amsinha@redhat.com>

amartyasinha force-pushed the restructure-template branch from 06fa137 to defa882 Compare April 21, 2026 08:04


		replace k8s.io/kube-openapi => k8s.io/kube-openapi v0.0.0-20250627150254-e9823e99808e //allow-merging

		replace github.com/openstack-k8s-operators/lib-common/modules/common => github.com/amartyasinha/lib-common/modules/common v0.0.0-20260416075014-135bf0acdeaf

Conversation

amartyasinha commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

openshift-ci Bot commented Apr 9, 2026

Uh oh!

softwarefactory-project-zuul Bot commented Apr 9, 2026

Uh oh!

softwarefactory-project-zuul Bot commented Apr 10, 2026

Uh oh!

bogdando commented Apr 10, 2026

Uh oh!

amartyasinha commented Apr 13, 2026

Uh oh!

gibizer left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

openshift-ci Bot commented Apr 16, 2026

Uh oh!

softwarefactory-project-zuul Bot commented Apr 16, 2026

Uh oh!

softwarefactory-project-zuul Bot commented Apr 16, 2026

Uh oh!

amartyasinha commented Apr 16, 2026

Uh oh!

gibizer left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

softwarefactory-project-zuul Bot commented Apr 16, 2026

Uh oh!

amartyasinha commented Apr 17, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

amartyasinha commented Apr 17, 2026

Uh oh!

softwarefactory-project-zuul Bot commented Apr 17, 2026

Uh oh!

stuggi commented Apr 17, 2026

Uh oh!

amartyasinha commented Apr 20, 2026

Uh oh!

stuggi commented Apr 20, 2026

Uh oh!

amartyasinha commented Apr 20, 2026

Uh oh!

stuggi commented Apr 20, 2026

Uh oh!

amartyasinha commented Apr 20, 2026

Uh oh!

amartyasinha commented Apr 21, 2026

amartyasinha commented Apr 9, 2026 •

edited

Loading