diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2ddb49d00..8f94c487d 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,47 @@
# Changelog
+## 2026-07-03 — [#1009](https://github.com/cobaltcore-dev/cortex/pull/1009)
+
+### cortex v0.2.2 (sha-f4d7df89)
+
+Non-breaking changes:
+- Include host information in error logs when multicluster connection fails ([#997](https://github.com/cobaltcore-dev/cortex/pull/997))
+- Fix Perses config for local development ([#1007](https://github.com/cobaltcore-dev/cortex/pull/1007))
+- Update `github.com/sapcc/go-bits` ([#1005](https://github.com/cobaltcore-dev/cortex/pull/1005))
+- Update kube-prometheus-stack Docker tag to v87.4.0 ([#998](https://github.com/cobaltcore-dev/cortex/pull/998))
+
+### cortex-shim v0.1.5 (sha-c98ddb05)
+
+Includes updated image sha-c98ddb05.
+
+### cortex-nova v0.0.79
+
+Includes updated charts cortex v0.2.2, cortex-postgres v0.6.8.
+
+### cortex-cinder v0.0.79
+
+Includes updated charts cortex v0.2.2, cortex-postgres v0.6.8.
+
+### cortex-manila v0.0.79
+
+Includes updated charts cortex v0.2.2, cortex-postgres v0.6.8.
+
+### cortex-crds v0.0.79
+
+Includes updated chart cortex v0.2.2.
+
+### cortex-ironcore v0.0.79
+
+Includes updated chart cortex v0.2.2.
+
+### cortex-pods v0.0.79
+
+Includes updated chart cortex v0.2.2.
+
+### cortex-placement-shim v0.1.5
+
+Includes updated chart cortex-shim v0.1.5.
+
## 2026-07-01 — [#1001](https://github.com/cobaltcore-dev/cortex/pull/1001)
### cortex v0.2.1 (sha-44b8aab7)
diff --git a/api/v1alpha1/datasource_types.go b/api/v1alpha1/datasource_types.go
index f9963a35c..fff321c48 100644
--- a/api/v1alpha1/datasource_types.go
+++ b/api/v1alpha1/datasource_types.go
@@ -52,7 +52,6 @@ const (
NovaDatasourceTypeFlavors NovaDatasourceType = "flavors"
NovaDatasourceTypeMigrations NovaDatasourceType = "migrations"
NovaDatasourceTypeAggregates NovaDatasourceType = "aggregates"
- NovaDatasourceTypeImages NovaDatasourceType = "images"
)
type NovaDatasource struct {
diff --git a/docs/reservations/failover-reservations.md b/docs/reservations/failover-reservations.md
index c2af22a8e..cd642d46c 100644
--- a/docs/reservations/failover-reservations.md
+++ b/docs/reservations/failover-reservations.md
@@ -26,15 +26,36 @@ The controller has two reconciliation modes:
```mermaid
flowchart TD
P1[List Hypervisors from K8s]
+ P1b["Build active-VM set from
Hypervisor CRD Status.Instances"]
P2["List VMs from Postgres
(vm_source.go)"]
- P3["Remove Invalid VMs from reservations
(e.g., vm:host mapping wrong or vm deleted)"]
+ P3["Remove Invalid VMs from reservations
(e.g., vm:host mapping wrong or vm deleted)
with postgres data-loss safeguard"]
P4["Remove Non-eligible VMs from reservations
(via eligibility rules, reservation_eligibility.go)"]
P5[Delete Empty Reservations]
P6["Create/Assign Reservations
(reservation_scheduling.go)"]
- P1 --> P2 --> P3 --> P4 --> P5 --> P6
+ P1 --> P1b --> P2 --> P3 --> P4 --> P5 --> P6
```
+#### Postgres Data-Loss Safeguard
+
+Before removing a VM from a failover reservation because it is missing from the
+postgres-derived VM source, the controller cross-checks the Hypervisor CRD
+`Status.Instances`. If the VM is still reported as active on any hypervisor, its
+allocation is preserved in the reservation.
+
+This safeguard prevents a postgres data loss or restore event from cascading into
+the mass deletion of all failover reservations. Without it, a wiped or partially
+restored Nova database would make every VM appear "deleted," causing the
+controller to empty and then garbage-collect all reservations -- leaving the
+entire fleet without failover coverage until postgres recovers and the
+reservations are rebuilt.
+
+The active-VM set (`vmsOnHypervisor`) is built once per reconciliation cycle by
+iterating over all Hypervisor CRD `Status.Instances` entries that are marked
+active. During the "Remove Invalid VMs" step, if a VM UUID is absent from the
+postgres VM list but present in this set, the allocation is kept and a log
+message is emitted.
+
### Watch-based Reconciliation
diff --git a/go.mod b/go.mod
index 97ef930e1..eff248f05 100644
--- a/go.mod
+++ b/go.mod
@@ -9,7 +9,7 @@ require (
github.com/ironcore-dev/ironcore v0.4.2
github.com/prometheus/client_golang v1.23.2
github.com/prometheus/client_model v0.6.2
- github.com/sapcc/go-bits v0.0.0-20260626143732-3999ce9f8fdc
+ github.com/sapcc/go-bits v0.0.0-20260703101005-27edea94477c
go.xyrillian.de/gg v1.11.1
k8s.io/api v0.36.2
k8s.io/apimachinery v0.36.2
diff --git a/go.sum b/go.sum
index 8941058a5..111d07a4f 100644
--- a/go.sum
+++ b/go.sum
@@ -204,8 +204,8 @@ github.com/rogpeppe/go-internal v1.14.1/go.mod h1:MaRKkUm5W0goXpeCfT7UZI6fk/L7L7
github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
github.com/sapcc/go-api-declarations v1.24.0 h1:sGBvOMVSM1olJlyvNoQSk7NX5uatXHKkztGDBPnTWMs=
github.com/sapcc/go-api-declarations v1.24.0/go.mod h1:ZWRTijvgF8o8aHg5stgg7u4DF6jFrd0X97le/uGlZsA=
-github.com/sapcc/go-bits v0.0.0-20260626143732-3999ce9f8fdc h1:oflzbbZ1C2tc0EMCwf5tETd6nlqA5TMnckihJqz7joU=
-github.com/sapcc/go-bits v0.0.0-20260626143732-3999ce9f8fdc/go.mod h1:WIr0qdewi9zEDh4egX9c7zB4e6JwrzTS3AA02y0EVxk=
+github.com/sapcc/go-bits v0.0.0-20260703101005-27edea94477c h1:jqVSQNGIZ+V7mP+bjKoz/OEVy5qWoGvaYCfwXfohEWw=
+github.com/sapcc/go-bits v0.0.0-20260703101005-27edea94477c/go.mod h1:uchHF8FNxncTCoQYDjGSO86AuxSzgIbN5mQzM95RYOA=
github.com/sirupsen/logrus v1.9.3 h1:dueUQJ1C2q9oE3F7wvmSGAaVtTmUizReu6fjN8uqzbQ=
github.com/sirupsen/logrus v1.9.3/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
github.com/spf13/cobra v1.10.2 h1:DMTTonx5m65Ic0GOoRY2c16WCbHxOOw6xxezuLaBpcU=
diff --git a/helm/bundles/cortex-cinder/Chart.yaml b/helm/bundles/cortex-cinder/Chart.yaml
index dfaa6cdce..bec946fca 100644
--- a/helm/bundles/cortex-cinder/Chart.yaml
+++ b/helm/bundles/cortex-cinder/Chart.yaml
@@ -5,23 +5,23 @@ apiVersion: v2
name: cortex-cinder
description: A Helm chart deploying Cortex for Cinder.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex-postgres
- name: cortex-postgres
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.6.7
+ version: 0.6.8
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-knowledge-controllers
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-scheduling-controllers
# Owner info adds a configmap to the kubernetes cluster with information on
diff --git a/helm/bundles/cortex-crds/Chart.yaml b/helm/bundles/cortex-crds/Chart.yaml
index 8f6b4cda1..b0402277a 100644
--- a/helm/bundles/cortex-crds/Chart.yaml
+++ b/helm/bundles/cortex-crds/Chart.yaml
@@ -5,13 +5,13 @@ apiVersion: v2
name: cortex-crds
description: A Helm chart deploying Cortex CRDs.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
# Owner info adds a configmap to the kubernetes cluster with information on
# the service owner. This makes it easier to find out who to contact in case
diff --git a/helm/bundles/cortex-ironcore/Chart.yaml b/helm/bundles/cortex-ironcore/Chart.yaml
index 762ee557c..0b91af43e 100644
--- a/helm/bundles/cortex-ironcore/Chart.yaml
+++ b/helm/bundles/cortex-ironcore/Chart.yaml
@@ -5,13 +5,13 @@ apiVersion: v2
name: cortex-ironcore
description: A Helm chart deploying Cortex for IronCore.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
# Owner info adds a configmap to the kubernetes cluster with information on
# the service owner. This makes it easier to find out who to contact in case
diff --git a/helm/bundles/cortex-manila/Chart.yaml b/helm/bundles/cortex-manila/Chart.yaml
index 76be70c05..b1f69a0d2 100644
--- a/helm/bundles/cortex-manila/Chart.yaml
+++ b/helm/bundles/cortex-manila/Chart.yaml
@@ -5,23 +5,23 @@ apiVersion: v2
name: cortex-manila
description: A Helm chart deploying Cortex for Manila.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex-postgres
- name: cortex-postgres
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.6.7
+ version: 0.6.8
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-knowledge-controllers
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-scheduling-controllers
# Owner info adds a configmap to the kubernetes cluster with information on
diff --git a/helm/bundles/cortex-nova/Chart.yaml b/helm/bundles/cortex-nova/Chart.yaml
index b803122b4..ebd6c4de4 100644
--- a/helm/bundles/cortex-nova/Chart.yaml
+++ b/helm/bundles/cortex-nova/Chart.yaml
@@ -5,23 +5,23 @@ apiVersion: v2
name: cortex-nova
description: A Helm chart deploying Cortex for Nova.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex-postgres
- name: cortex-postgres
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.6.7
+ version: 0.6.8
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-knowledge-controllers
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
alias: cortex-scheduling-controllers
# Owner info adds a configmap to the kubernetes cluster with information on
diff --git a/helm/bundles/cortex-nova/templates/datasources.yaml b/helm/bundles/cortex-nova/templates/datasources.yaml
index 582effac2..f9160602f 100644
--- a/helm/bundles/cortex-nova/templates/datasources.yaml
+++ b/helm/bundles/cortex-nova/templates/datasources.yaml
@@ -337,30 +337,6 @@ spec:
---
apiVersion: cortex.cloud/v1alpha1
kind: Datasource
-metadata:
- name: nova-images
-spec:
- schedulingDomain: nova
- databaseSecretRef:
- name: cortex-nova-postgres
- namespace: {{ .Release.Namespace }}
- {{- if .Values.openstack.sso.enabled }}
- ssoSecretRef:
- name: cortex-nova-openstack-sso
- namespace: {{ .Release.Namespace }}
- {{- end }}
- type: openstack
- openstack:
- syncInterval: 3600s
- secretRef:
- name: cortex-nova-openstack-keystone
- namespace: {{ .Release.Namespace }}
- type: nova
- nova:
- type: images
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Datasource
metadata:
name: limes-project-commitments
spec:
diff --git a/helm/bundles/cortex-placement-shim/Chart.yaml b/helm/bundles/cortex-placement-shim/Chart.yaml
index 95ecb8cec..07e09aab7 100644
--- a/helm/bundles/cortex-placement-shim/Chart.yaml
+++ b/helm/bundles/cortex-placement-shim/Chart.yaml
@@ -5,13 +5,13 @@ apiVersion: v2
name: cortex-placement-shim
description: A Helm chart deploying the Cortex placement shim.
type: application
-version: 0.1.4
+version: 0.1.5
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex-shim
- name: cortex-shim
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.1.4
+ version: 0.1.5
# Owner info adds a configmap to the kubernetes cluster with information on
# the service owner. This makes it easier to find out who to contact in case
# of issues. See: https://github.com/sapcc/helm-charts/pkgs/container/helm-charts%2Fowner-info
diff --git a/helm/bundles/cortex-pods/Chart.yaml b/helm/bundles/cortex-pods/Chart.yaml
index 6a773d921..6bbcf92ce 100644
--- a/helm/bundles/cortex-pods/Chart.yaml
+++ b/helm/bundles/cortex-pods/Chart.yaml
@@ -5,13 +5,13 @@ apiVersion: v2
name: cortex-pods
description: A Helm chart deploying Cortex for Pods.
type: application
-version: 0.0.78
+version: 0.0.79
appVersion: 0.1.0
dependencies:
# from: file://../../library/cortex
- name: cortex
repository: oci://ghcr.io/cobaltcore-dev/cortex/charts
- version: 0.2.1
+ version: 0.2.2
# Owner info adds a configmap to the kubernetes cluster with information on
# the service owner. This makes it easier to find out who to contact in case
diff --git a/helm/dev/cortex-prometheus-operator/Chart.yaml b/helm/dev/cortex-prometheus-operator/Chart.yaml
index 59859b611..7e2d4dbba 100644
--- a/helm/dev/cortex-prometheus-operator/Chart.yaml
+++ b/helm/dev/cortex-prometheus-operator/Chart.yaml
@@ -10,4 +10,4 @@ dependencies:
# CRDs of the prometheus operator, such as PrometheusRule, ServiceMonitor, etc.
- name: kube-prometheus-stack
repository: oci://ghcr.io/prometheus-community/charts
- version: 87.3.0
+ version: 87.10.1
diff --git a/helm/library/cortex-postgres/Chart.yaml b/helm/library/cortex-postgres/Chart.yaml
index 207c80572..b64823e29 100644
--- a/helm/library/cortex-postgres/Chart.yaml
+++ b/helm/library/cortex-postgres/Chart.yaml
@@ -5,5 +5,5 @@ apiVersion: v2
name: cortex-postgres
description: Postgres setup for Cortex.
type: application
-version: 0.6.7
+version: 0.6.8
appVersion: "sha-af707446"
diff --git a/helm/library/cortex-shim/Chart.yaml b/helm/library/cortex-shim/Chart.yaml
index f221a68b6..0e6b0ddfb 100644
--- a/helm/library/cortex-shim/Chart.yaml
+++ b/helm/library/cortex-shim/Chart.yaml
@@ -2,7 +2,7 @@ apiVersion: v2
name: cortex-shim
description: A Helm chart to distribute cortex shims.
type: application
-version: 0.1.4
-appVersion: "sha-44b8aab7"
+version: 0.1.5
+appVersion: "sha-c8295850"
icon: "https://example.com/icon.png"
dependencies: []
diff --git a/helm/library/cortex/Chart.yaml b/helm/library/cortex/Chart.yaml
index 5d1b9bdef..b746c98f3 100644
--- a/helm/library/cortex/Chart.yaml
+++ b/helm/library/cortex/Chart.yaml
@@ -2,7 +2,7 @@ apiVersion: v2
name: cortex
description: A Helm chart to distribute cortex.
type: application
-version: 0.2.1
-appVersion: "sha-44b8aab7"
+version: 0.2.2
+appVersion: "sha-8d888397"
icon: "https://example.com/icon.png"
dependencies: []
diff --git a/internal/knowledge/datasources/plugins/openstack/controller_test.go b/internal/knowledge/datasources/plugins/openstack/controller_test.go
index 586238d54..899e83237 100644
--- a/internal/knowledge/datasources/plugins/openstack/controller_test.go
+++ b/internal/knowledge/datasources/plugins/openstack/controller_test.go
@@ -104,7 +104,6 @@ func TestNovaDatasourceTypeConstants(t *testing.T) {
{v1alpha1.NovaDatasourceTypeFlavors, "flavors"},
{v1alpha1.NovaDatasourceTypeMigrations, "migrations"},
{v1alpha1.NovaDatasourceTypeAggregates, "aggregates"},
- {v1alpha1.NovaDatasourceTypeImages, "images"},
}
for _, test := range tests {
diff --git a/internal/knowledge/datasources/plugins/openstack/nova/nova_api.go b/internal/knowledge/datasources/plugins/openstack/nova/nova_api.go
index 906a46a1e..b7e25c2ba 100644
--- a/internal/knowledge/datasources/plugins/openstack/nova/nova_api.go
+++ b/internal/knowledge/datasources/plugins/openstack/nova/nova_api.go
@@ -10,18 +10,15 @@ import (
"log/slog"
"net/http"
"net/url"
- "strings"
"time"
"github.com/cobaltcore-dev/cortex/api/v1alpha1"
"github.com/cobaltcore-dev/cortex/internal/knowledge/datasources"
"github.com/cobaltcore-dev/cortex/pkg/keystone"
"github.com/gophercloud/gophercloud/v2"
- "github.com/gophercloud/gophercloud/v2/openstack"
"github.com/gophercloud/gophercloud/v2/openstack/compute/v2/aggregates"
"github.com/gophercloud/gophercloud/v2/openstack/compute/v2/flavors"
"github.com/gophercloud/gophercloud/v2/openstack/compute/v2/servers"
- glanceimages "github.com/gophercloud/gophercloud/v2/openstack/image/v2/images"
"github.com/gophercloud/gophercloud/v2/pagination"
"github.com/prometheus/client_golang/prometheus"
"github.com/sapcc/go-bits/liquidapi"
@@ -43,8 +40,6 @@ type NovaAPI interface {
GetAllMigrations(ctx context.Context) ([]Migration, error)
// Get all aggregates.
GetAllAggregates(ctx context.Context) ([]Aggregate, error)
- // Get all Glance images with pre-computed os_type.
- GetAllImages(ctx context.Context) ([]Image, error)
}
// API for OpenStack Nova.
@@ -57,8 +52,6 @@ type novaAPI struct {
conf v1alpha1.NovaDatasource
// Authenticated OpenStack compute service client.
sc *gophercloud.ServiceClient
- // Authenticated Glance image service client (only used for NovaDatasourceTypeImages).
- glance *gophercloud.ServiceClient
// OS type prober for determining VM operating system type (only for NovaDatasourceTypeServers).
osTypeProber *liquidapi.OSTypeProber
}
@@ -90,16 +83,6 @@ func (api *novaAPI) Init(ctx context.Context) error {
// Since 2.61, the extra_specs are returned in the flavor details.
Microversion: "2.61",
}
- // Initialize the Glance client only when this datasource is used for images.
- if api.conf.Type == v1alpha1.NovaDatasourceTypeImages {
- glanceClient, err := openstack.NewImageV2(provider, gophercloud.EndpointOpts{
- Availability: gophercloud.Availability(sameAsKeystone),
- })
- if err != nil {
- return fmt.Errorf("failed to create Glance client: %w", err)
- }
- api.glance = glanceClient
- }
// Initialize the OS type prober only for the servers datasource.
if api.conf.Type == v1alpha1.NovaDatasourceTypeServers {
eo := gophercloud.EndpointOpts{Availability: gophercloud.Availability(sameAsKeystone)}
@@ -510,75 +493,6 @@ func (api *novaAPI) GetAllAggregates(ctx context.Context) ([]Aggregate, error) {
return aggregates, nil
}
-// GetAllImages fetches all Glance images and returns them with pre-computed os_type.
-// See deriveOSType for the derivation logic.
-func (api *novaAPI) GetAllImages(ctx context.Context) ([]Image, error) {
- if api.glance == nil {
- return nil, fmt.Errorf("glance client not initialized: datasource type must be %q", v1alpha1.NovaDatasourceTypeImages)
- }
-
- label := Image{}.TableName()
- slog.Info("fetching nova data", "label", label)
- if api.mon.RequestTimer != nil {
- hist := api.mon.RequestTimer.WithLabelValues(label)
- timer := prometheus.NewTimer(hist)
- defer timer.ObserveDuration()
- }
-
- var result []Image
- opts := glanceimages.ListOpts{Limit: 1000}
- err := glanceimages.List(api.glance, opts).EachPage(ctx, func(_ context.Context, page pagination.Page) (bool, error) {
- imgs, err := glanceimages.ExtractImages(page)
- if err != nil {
- return false, err
- }
- for _, img := range imgs {
- result = append(result, Image{
- ID: img.ID,
- OSType: deriveOSType(img.Properties, img.Tags),
- })
- }
- return true, nil
- })
- if err != nil {
- return nil, fmt.Errorf("failed to list Glance images: %w", err)
- }
- slog.Info("fetched", "label", label, "count", len(result))
- return result, nil
-}
-
-// deriveOSType computes os_type from image properties and tags.
-// Mirrors the logic of OSTypeProber.findFromImage in github.com/sapcc/go-bits/liquidapi,
-// with two intentional simplifications:
-// 1. No regex validation on vmware_ostype — Nova validates that field at VM boot time,
-// so any value stored in Glance is already valid.
-// 2. Volume-booted VMs are not yet supported — os_type will be "unknown" for them.
-// Supporting them would require per-VM Cinder calls (volume_image_metadata.vmware_ostype)
-// either at server sync time or via a dedicated datasource.
-func deriveOSType(properties map[string]any, tags []string) string {
- if v, ok := properties["vmware_ostype"]; ok {
- if s, ok := v.(string); ok && s != "" {
- return s
- }
- }
- var osType string
- for _, tag := range tags {
- if after, ok := strings.CutPrefix(tag, "ostype:"); ok {
- if osType == "" {
- osType = after
- } else {
- // multiple ostype: tags → ambiguous, fall through to unknown
- osType = ""
- break
- }
- }
- }
- if osType != "" {
- return osType
- }
- return "unknown"
-}
-
// initOSTypeProber safely creates an OSTypeProber, returning nil on any error or panic.
func initOSTypeProber(provider *gophercloud.ProviderClient, eo gophercloud.EndpointOpts) (prober *liquidapi.OSTypeProber) {
defer func() {
diff --git a/internal/knowledge/datasources/plugins/openstack/nova/nova_api_test.go b/internal/knowledge/datasources/plugins/openstack/nova/nova_api_test.go
index 63f83c176..49f0af4b4 100644
--- a/internal/knowledge/datasources/plugins/openstack/nova/nova_api_test.go
+++ b/internal/knowledge/datasources/plugins/openstack/nova/nova_api_test.go
@@ -538,66 +538,6 @@ func TestNovaAPI_GetAllHypervisors_DeduplicatesHypervisors(t *testing.T) {
}
}
-func TestDeriveOSType(t *testing.T) {
- tests := []struct {
- name string
- properties map[string]any
- tags []string
- want string
- }{
- {
- name: "vmware_ostype property wins",
- properties: map[string]any{"vmware_ostype": "windows8Server64Guest"},
- tags: []string{"ostype:linux"},
- want: "windows8Server64Guest",
- },
- {
- name: "vmware_ostype empty string falls through to tags",
- properties: map[string]any{"vmware_ostype": ""},
- tags: []string{"ostype:debian"},
- want: "debian",
- },
- {
- name: "vmware_ostype not a string falls through",
- properties: map[string]any{"vmware_ostype": 42},
- tags: []string{"ostype:centos"},
- want: "centos",
- },
- {
- name: "single ostype tag",
- properties: map[string]any{},
- tags: []string{"ostype:ubuntu", "env:prod"},
- want: "ubuntu",
- },
- {
- name: "multiple ostype tags: ambiguous, returns unknown",
- properties: map[string]any{},
- tags: []string{"ostype:ubuntu", "ostype:debian"},
- want: "unknown",
- },
- {
- name: "no properties, no tags",
- properties: map[string]any{},
- tags: nil,
- want: "unknown",
- },
- {
- name: "tags without ostype prefix",
- properties: map[string]any{},
- tags: []string{"env:prod", "region:eu"},
- want: "unknown",
- },
- }
-
- for _, tt := range tests {
- t.Run(tt.name, func(t *testing.T) {
- if got := deriveOSType(tt.properties, tt.tags); got != tt.want {
- t.Errorf("deriveOSType() = %q, want %q", got, tt.want)
- }
- })
- }
-}
-
func TestNovaAPI_GetAllMigrations_DeduplicatesMigrations(t *testing.T) {
tests := []struct {
name string
diff --git a/internal/knowledge/datasources/plugins/openstack/nova/nova_sync.go b/internal/knowledge/datasources/plugins/openstack/nova/nova_sync.go
index ec0533f84..a47449c27 100644
--- a/internal/knowledge/datasources/plugins/openstack/nova/nova_sync.go
+++ b/internal/knowledge/datasources/plugins/openstack/nova/nova_sync.go
@@ -46,8 +46,6 @@ func (s *NovaSyncer) Init(ctx context.Context) error {
tables = append(tables, s.DB.AddTable(Migration{}))
case v1alpha1.NovaDatasourceTypeAggregates:
tables = append(tables, s.DB.AddTable(Aggregate{}))
- case v1alpha1.NovaDatasourceTypeImages:
- tables = append(tables, s.DB.AddTable(Image{}))
}
return s.DB.CreateTable(tables...)
}
@@ -70,8 +68,6 @@ func (s *NovaSyncer) Sync(ctx context.Context) (int64, error) {
nResults, err = s.SyncAllMigrations(ctx)
case v1alpha1.NovaDatasourceTypeAggregates:
nResults, err = s.SyncAllAggregates(ctx)
- case v1alpha1.NovaDatasourceTypeImages:
- nResults, err = s.SyncAllImages(ctx)
}
return nResults, err
}
@@ -202,26 +198,6 @@ func (s *NovaSyncer) SyncAllMigrations(ctx context.Context) (int64, error) {
return int64(len(allMigrations)), nil
}
-// Sync all Glance images into the database with pre-computed os_type.
-func (s *NovaSyncer) SyncAllImages(ctx context.Context) (int64, error) {
- allImages, err := s.API.GetAllImages(ctx)
- if err != nil {
- return 0, err
- }
- err = db.ReplaceAll(s.DB, allImages...)
- if err != nil {
- return 0, err
- }
- label := Image{}.TableName()
- if s.Mon.ObjectsGauge != nil {
- s.Mon.ObjectsGauge.WithLabelValues(label).Set(float64(len(allImages)))
- }
- if s.Mon.RequestProcessedCounter != nil {
- s.Mon.RequestProcessedCounter.WithLabelValues(label).Inc()
- }
- return int64(len(allImages)), nil
-}
-
// Sync the OpenStack aggregates into the database.
func (s *NovaSyncer) SyncAllAggregates(ctx context.Context) (int64, error) {
allAggregates, err := s.API.GetAllAggregates(ctx)
diff --git a/internal/knowledge/datasources/plugins/openstack/nova/nova_sync_test.go b/internal/knowledge/datasources/plugins/openstack/nova/nova_sync_test.go
index fdd826171..9761b9a93 100644
--- a/internal/knowledge/datasources/plugins/openstack/nova/nova_sync_test.go
+++ b/internal/knowledge/datasources/plugins/openstack/nova/nova_sync_test.go
@@ -55,10 +55,6 @@ func (m *mockNovaAPI) GetAllAggregates(ctx context.Context) ([]Aggregate, error)
return []Aggregate{{Name: "aggregate1"}}, nil
}
-func (m *mockNovaAPI) GetAllImages(ctx context.Context) ([]Image, error) {
- return []Image{{ID: "img-1", OSType: "windows8Server64Guest"}}, nil
-}
-
func TestNovaSyncer_Init(t *testing.T) {
dbEnv := testlibDB.SetupDBEnv(t)
testDB := db.DB{DbMap: dbEnv.DbMap}
@@ -271,35 +267,3 @@ func TestNovaSyncer_SyncAggregates(t *testing.T) {
t.Fatalf("expected 1 aggregate, got %d", n)
}
}
-
-func TestNovaSyncer_SyncImages(t *testing.T) {
- dbEnv := testlibDB.SetupDBEnv(t)
- testDB := db.DB{DbMap: dbEnv.DbMap}
- defer dbEnv.Close()
- mon := datasources.Monitor{}
- syncer := &NovaSyncer{
- DB: testDB,
- Mon: mon,
- Conf: v1alpha1.NovaDatasource{Type: v1alpha1.NovaDatasourceTypeImages},
- API: &mockNovaAPI{},
- }
-
- ctx := t.Context()
- if err := syncer.Init(ctx); err != nil {
- t.Fatalf("failed to init images syncer: %v", err)
- }
- n, err := syncer.Sync(ctx)
- if err != nil {
- t.Fatalf("expected no error, got %v", err)
- }
- if n != 1 {
- t.Fatalf("expected 1 image, got %d", n)
- }
- var images []Image
- if _, err := testDB.Select(&images, "SELECT * FROM "+Image{}.TableName()); err != nil {
- t.Fatalf("select images: %v", err)
- }
- if len(images) != 1 || images[0].ID != "img-1" || images[0].OSType != "windows8Server64Guest" {
- t.Errorf("unexpected images in DB: %+v", images)
- }
-}
diff --git a/internal/knowledge/datasources/plugins/openstack/nova/nova_types.go b/internal/knowledge/datasources/plugins/openstack/nova/nova_types.go
index 17c422194..544b74572 100644
--- a/internal/knowledge/datasources/plugins/openstack/nova/nova_types.go
+++ b/internal/knowledge/datasources/plugins/openstack/nova/nova_types.go
@@ -512,17 +512,3 @@ func (Aggregate) TableName() string { return "openstack_aggregates_v2" }
// Index for the openstack model.
func (Aggregate) Indexes() map[string][]string { return nil }
-
-// Image stores pre-computed os_type for a Glance image UUID.
-// Populated by the NovaDatasourceTypeImages syncer from the Glance API.
-// Used by the CR usage API to include os_type in VM subresources without live API calls.
-type Image struct {
- ID string `json:"id" db:"id,primarykey"`
- OSType string `json:"os_type" db:"os_type"`
-}
-
-// Table in which the openstack model is stored.
-func (Image) TableName() string { return "openstack_images" }
-
-// Index for the openstack model.
-func (Image) Indexes() map[string][]string { return nil }
diff --git a/internal/scheduling/reservations/capacity/controller.go b/internal/scheduling/reservations/capacity/controller.go
index fd71f5a05..1e69b22cb 100644
--- a/internal/scheduling/reservations/capacity/controller.go
+++ b/internal/scheduling/reservations/capacity/controller.go
@@ -150,6 +150,16 @@ func (c *Controller) computeVMUsage(
return result
}
+ // Pre-populate all (flavorGroup, AZ) pairs with fresh=true and zero resources.
+ // This ensures that groups with no running VMs still get fresh=true so that
+ // RunningInstances/RunningResources are correctly zeroed out in the CRD.
+ azs := availabilityZones(hvs)
+ for groupName := range flavorGroups {
+ for _, az := range azs {
+ result[vmUsageKey{group: groupName, az: az}] = vmUsage{fresh: true}
+ }
+ }
+
flavorToGroup := make(map[string]string)
flavorMemBytes := make(map[string]int64)
flavorVCPUs := make(map[string]int64)
diff --git a/internal/scheduling/reservations/capacity/controller_test.go b/internal/scheduling/reservations/capacity/controller_test.go
index fe026416b..c33d98be4 100644
--- a/internal/scheduling/reservations/capacity/controller_test.go
+++ b/internal/scheduling/reservations/capacity/controller_test.go
@@ -678,6 +678,135 @@ func TestSumCommittedCapacity(t *testing.T) {
// TestProbeScheduler_SubtractsReservationBlocksWhenNotIgnored verifies that placeable-probe
// slot counting subtracts per-host reservation blocks in addition to hv.Status.Allocation.
+// mockVMSource is a test implementation of VMSource that returns a fixed list of VMs.
+type mockVMSource struct {
+ vms []reservations.VM
+ err error
+}
+
+func (m *mockVMSource) ListVMs(_ context.Context) ([]reservations.VM, error) {
+ return m.vms, m.err
+}
+
+func (m *mockVMSource) ListVMsByProject(_ context.Context, _ string) ([]reservations.VM, error) {
+ return m.vms, m.err
+}
+
+func (m *mockVMSource) ListVMsOnHypervisors(_ context.Context, _ *hv1.HypervisorList, _ bool) ([]reservations.VM, error) {
+ return m.vms, m.err
+}
+
+func (m *mockVMSource) GetVM(_ context.Context, _ string) (*reservations.VM, error) {
+ return nil, nil
+}
+
+func (m *mockVMSource) IsServerActive(_ context.Context, _ string) (bool, error) {
+ return false, nil
+}
+
+func (m *mockVMSource) GetDeletedVMInfo(_ context.Context, _ string) (*reservations.DeletedVMInfo, error) {
+ return nil, nil
+}
+
+// TestComputeVMUsage_ZerosOutWhenAllVMsRemoved verifies that after a successful
+// ListVMsOnHypervisors call that returns no VMs for a (flavorGroup, AZ) pair,
+// the result map entry has fresh=true with zero instances/resources, ensuring
+// RunningInstances and RunningResources are zeroed out in the CRD.
+func TestComputeVMUsage_ZerosOutWhenAllVMsRemoved(t *testing.T) {
+ const (
+ groupName = "hana-v2"
+ az = "qa-de-1a"
+ memMB = 4096
+ memBytes = int64(memMB) * 1024 * 1024
+ )
+
+ scheme := newTestScheme(t)
+ hv := newHypervisor("host-1", az, memBytes, "vm1")
+ knowledge := newFlavorGroupKnowledge(t, groupName, memMB)
+
+ // Pre-create CRD with non-zero RunningInstances to simulate prior state.
+ crdName := crdNameFor(groupName, az)
+ existing := &v1alpha1.FlavorGroupCapacity{
+ ObjectMeta: metav1.ObjectMeta{Name: crdName},
+ Spec: v1alpha1.FlavorGroupCapacitySpec{
+ FlavorGroup: groupName,
+ AvailabilityZone: az,
+ },
+ }
+
+ fakeClient := fake.NewClientBuilder().
+ WithScheme(scheme).
+ WithObjects(knowledge, hv, existing).
+ WithStatusSubresource(&v1alpha1.FlavorGroupCapacity{}, &v1alpha1.Knowledge{}).
+ Build()
+
+ // Set RunningInstances to non-zero via a status patch.
+ patch := client.MergeFrom(existing.DeepCopy())
+ existing.Status.RunningInstances = 5
+ existing.Status.RunningResources = map[string]resource.Quantity{
+ string(v1alpha1.CommittedResourceTypeMemory): *resource.NewQuantity(memBytes*5, resource.BinarySI),
+ }
+ if err := fakeClient.Status().Patch(context.Background(), existing, patch); err != nil {
+ t.Fatalf("failed to patch CRD status: %v", err)
+ }
+
+ // Scheduler returns host-1 so probes succeed.
+ schedulerServer := newMockSchedulerServer(t, []string{"host-1"})
+ defer schedulerServer.Close()
+
+ // VMSource returns empty list (all VMs removed).
+ vmSource := &mockVMSource{vms: []reservations.VM{}}
+
+ ctrl := NewController(fakeClient, Config{
+ SchedulerURL: schedulerServer.URL,
+ TotalPipeline: "kvm-report-capacity",
+ PlaceablePipeline: "kvm-general-purpose",
+ }, vmSource)
+
+ smallFlavor := compute.FlavorInGroup{Name: groupName + "-small", MemoryMB: memMB, VCPUs: 2}
+ groupData := compute.FlavorGroupFeature{
+ Name: groupName,
+ SmallestFlavor: smallFlavor,
+ Flavors: []compute.FlavorInGroup{smallFlavor},
+ }
+ hvByName := map[string]hv1.Hypervisor{"host-1": *hv}
+ groups := map[string]compute.FlavorGroupFeature{groupName: groupData}
+
+ // Compute VM usage — should return fresh=true with zero instances.
+ usageByKey := ctrl.computeVMUsage(context.Background(), groups, []hv1.Hypervisor{*hv})
+ key := vmUsageKey{group: groupName, az: az}
+ usage, exists := usageByKey[key]
+ if !exists {
+ t.Fatalf("expected usage entry for key %v, got none", key)
+ }
+ if !usage.fresh {
+ t.Errorf("usage.fresh = false, want true (successful call with no VMs)")
+ }
+ if usage.instances != 0 {
+ t.Errorf("usage.instances = %d, want 0", usage.instances)
+ }
+
+ // Now run reconcileAZ to verify the CRD gets zeroed out.
+ if err := ctrl.reconcileAZ(context.Background(), az, groups, hvByName, map[string]int64{}, usageByKey); err != nil {
+ t.Fatalf("reconcileAZ failed: %v", err)
+ }
+
+ var crd v1alpha1.FlavorGroupCapacity
+ if err := fakeClient.Get(context.Background(), types.NamespacedName{Name: crdName}, &crd); err != nil {
+ t.Fatalf("failed to get CRD: %v", err)
+ }
+ if crd.Status.RunningInstances != 0 {
+ t.Errorf("RunningInstances = %d, want 0 (all VMs removed)", crd.Status.RunningInstances)
+ }
+ if crd.Status.RunningResources != nil {
+ for k, v := range crd.Status.RunningResources {
+ if !v.IsZero() {
+ t.Errorf("RunningResources[%s] = %s, want 0", k, v.String())
+ }
+ }
+ }
+}
+
func TestProbeScheduler_SubtractsReservationBlocksWhenNotIgnored(t *testing.T) {
const memMB = 4096
const memBytes = int64(memMB) * 1024 * 1024
diff --git a/pkg/multicluster/client.go b/pkg/multicluster/client.go
index 6af557598..fcaecaf93 100644
--- a/pkg/multicluster/client.go
+++ b/pkg/multicluster/client.go
@@ -345,7 +345,8 @@ func (c *Client) Get(ctx context.Context, key client.ObjectKey, obj client.Objec
}
if !apierrors.IsNotFound(err) {
log.Error(err, "error checking for duplicate resource in cluster",
- "gvk", gvk, "namespace", key.Namespace, "name", key.Name)
+ "gvk", gvk, "namespace", key.Namespace, "name", key.Name,
+ "host", cl.GetConfig().Host)
}
continue
}
@@ -357,7 +358,8 @@ func (c *Client) Get(ctx context.Context, key client.ObjectKey, obj client.Objec
}
if !apierrors.IsNotFound(err) {
log.Error(err, "error getting resource from cluster", "gvk", gvk,
- "namespace", key.Namespace, "name", key.Name)
+ "namespace", key.Namespace, "name", key.Name,
+ "host", cl.GetConfig().Host)
}
}
if !found {
@@ -392,7 +394,8 @@ func (c *Client) List(ctx context.Context, list client.ObjectList, opts ...clien
for _, cl := range clusters {
listCopy := list.DeepCopyObject().(client.ObjectList)
if err := cl.GetClient().List(ctx, listCopy, opts...); err != nil {
- log.Error(err, "error listing resources from cluster", "gvk", gvk)
+ log.Error(err, "error listing resources from cluster",
+ "gvk", gvk, "host", cl.GetConfig().Host)
continue
}
items, err := meta.ExtractList(listCopy)
@@ -632,7 +635,7 @@ func (c *subResourceClient) Get(ctx context.Context, obj, subResource client.Obj
if !apierrors.IsNotFound(err) {
log.Error(err, "error checking for duplicate sub-resource in cluster",
"gvk", gvk, "namespace", obj.GetNamespace(), "name", obj.GetName(),
- "subresource", c.subResource)
+ "subresource", c.subResource, "host", cl.GetConfig().Host)
}
continue
}
@@ -646,7 +649,7 @@ func (c *subResourceClient) Get(ctx context.Context, obj, subResource client.Obj
if !apierrors.IsNotFound(err) {
log.Error(err, "error getting sub-resource from cluster", "gvk", gvk,
"namespace", obj.GetNamespace(), "name", obj.GetName(),
- "subresource", c.subResource)
+ "subresource", c.subResource, "host", cl.GetConfig().Host)
}
}
if !found {
diff --git a/tools/perses/values.yaml b/tools/perses/values.yaml
index ec489e2f4..30e68380f 100644
--- a/tools/perses/values.yaml
+++ b/tools/perses/values.yaml
@@ -4,12 +4,24 @@
sidecar:
enabled: true
+volumes:
+ - name: perses-plugins
+ emptyDir: {}
+
+volumeMounts:
+ - name: perses-plugins
+ mountPath: /etc/perses/plugins
+
config:
provisioning:
folders:
- /etc/perses/provisioning
- /etc/perses/datasources
interval: 1m
+ plugin:
+ path: /etc/perses/plugins
+ archive_paths:
+ - /etc/perses/plugins-archive
datasources:
- kind: GlobalDatasource