Skip to content

pkg/cvo/metrics: Add a cluster_version_capability metric #755

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.

Sign up for GitHub

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

Merged
openshift-merge-robot merged 1 commit into from
Mar 23, 2022
Merged

pkg/cvo/metrics: Add a cluster_version_capability metric #755

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 21 additions & 0 deletions pkg/cvo/metrics.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,6 +49,7 @@ type operatorMetrics struct {

version *prometheus.GaugeVec
availableUpdates *prometheus.GaugeVec
capability *prometheus.GaugeVec
clusterOperatorUp *prometheus.GaugeVec
clusterOperatorConditions *prometheus.GaugeVec
clusterOperatorConditionTransitions *prometheus.GaugeVec
Expand Down Expand Up @@ -87,6 +88,10 @@ version for 'cluster', or empty for 'initial'.
Name: "cluster_version_available_updates",
Help: "Report the count of available versions for an upstream and channel.",
}, []string{"upstream", "channel"}),
capability: prometheus.NewGaugeVec(prometheus.GaugeOpts{
Name: "cluster_version_capability",
Help: "Report currently enabled cluster capabilities. 0 is disabled, and 1 is enabled.",
}, []string{"name"}),
clusterOperatorUp: prometheus.NewGaugeVec(prometheus.GaugeOpts{
Name: "cluster_operator_up",
Help: "Reports key highlights of the active cluster operators.",
Expand Down Expand Up @@ -333,6 +338,7 @@ func (m *operatorMetrics) clusterOperatorChanged(oldObj, obj interface{}) {
func (m *operatorMetrics) Describe(ch chan<- *prometheus.Desc) {
ch <- m.version.WithLabelValues("", "", "", "").Desc()
ch <- m.availableUpdates.WithLabelValues("", "").Desc()
ch <- m.capability.WithLabelValues("").Desc()
ch <- m.clusterOperatorUp.WithLabelValues("", "").Desc()
ch <- m.clusterOperatorConditions.WithLabelValues("", "", "").Desc()
ch <- m.clusterOperatorConditionTransitions.WithLabelValues("", "").Desc()
Expand Down Expand Up @@ -432,6 +438,21 @@ func (m *operatorMetrics) Collect(ch chan<- prometheus.Metric) {
ch <- g
}

enabledCapabilities := make(map[configv1.ClusterVersionCapability]struct{}, len(cv.Status.Capabilities.EnabledCapabilities))
for _, capability := range cv.Status.Capabilities.EnabledCapabilities {
g := m.capability.WithLabelValues(string(capability))
g.Set(float64(1))
ch <- g
enabledCapabilities[capability] = struct{}{}
}
for _, capability := range cv.Status.Capabilities.KnownCapabilities {
Copy link
Member

@LalatenduMohanty LalatenduMohanty Mar 23, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I am not clear on why we need to send known capabilities.

Copy link
Member Author

@wking wking Mar 23, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Two use cases:

  • conditional updates using local PromQL can distinguish between "yeah, that cap is not enabled here" (value 0) and "err, something's broken with scraping this metrics" (no match)
  • Telemetry can gauge the install-fraction for a capability (countOfClustersWithCapEnabled / countOfClustersWithCapKnown). with count by (name) (cluster_version_capability == 1) / count by (name) (cluster_version_capability). If cluster_version_capability did not include known-but-not-enabled caps, you'd need some added filtering to get at countOfClustersWithCapKnown.

if _, ok := enabledCapabilities[capability]; !ok {
g := m.capability.WithLabelValues(string(capability))
g.Set(float64(0))
ch <- g
}
}

for _, condition := range cv.Status.Conditions {
if condition.Status != configv1.ConditionFalse && condition.Status != configv1.ConditionTrue {
klog.V(2).Infof("skipping metrics for ClusterVersion condition %s=%s (neither True nor False)", condition.Type, condition.Status)
Expand Down