1
0
Fork 0
mirror of https://github.com/kubernetes-sigs/node-feature-discovery.git synced 2025-03-28 10:47:23 +00:00

metrics: add nfd_node_update_failures_total counter

Add a new counter for tracking node update failures from nfd-master.
This tracks both normal feature updates and the --prune sub-command.
This is a simple counter without any additional labels - nfd-master logs
can be used for further diagnostics.
This commit is contained in:
Markus Lehtonen 2023-08-01 14:54:24 +03:00
parent 9ed191808d
commit b90f2c318e
4 changed files with 13 additions and 3 deletions

View file

@ -18,6 +18,7 @@ The exposed metrics are
| `nfd_master_build_info` | Gauge | Version from which nfd-master was built
| `nfd_worker_build_info` | Gauge | Version from which nfd-worker was built
| `nfd_node_updates_total` | Counter | Number of nodes updated
| `nfd_node_update_failures_total` | Counter | Number of nodes update failures
| `nfd_nodefeaturerule_processing_duration_seconds` | Histogram | Time taken to process NodeFeatureRule objects
| `nfd_feature_discovery_duration_seconds` | Histogram | Time taken to discover features on a node

View file

@ -28,9 +28,10 @@ import (
// When adding metric names, see https://prometheus.io/docs/practices/naming/#metric-names
const (
buildInfoQuery = "nfd_master_build_info"
nodeUpdatesQuery = "nfd_node_updates_total"
nfrProcessingTimeQuery = "nfd_nodefeaturerule_processing_duration_seconds"
buildInfoQuery = "nfd_master_build_info"
nodeUpdatesQuery = "nfd_node_updates_total"
nodeUpdateFailuresQuery = "nfd_node_update_failures_total"
nfrProcessingTimeQuery = "nfd_nodefeaturerule_processing_duration_seconds"
)
var (
@ -47,6 +48,10 @@ var (
Name: nodeUpdatesQuery,
Help: "Number of nodes updated by the master.",
})
nodeUpdateFailures = prometheus.NewCounter(prometheus.CounterOpts{
Name: nodeUpdateFailuresQuery,
Help: "Number of node update failures.",
})
nfrProcessingTime = prometheus.NewHistogramVec(
prometheus.HistogramOpts{
Name: nfrProcessingTimeQuery,
@ -70,6 +75,7 @@ func runMetricsServer(port int) {
r := prometheus.NewRegistry()
r.MustRegister(buildInfo,
nodeUpdates,
nodeUpdateFailures,
nfrProcessingTime)
mux := http.NewServeMux()

View file

@ -452,6 +452,7 @@ func (m *nfdMaster) prune() error {
// Prune labels and extended resources
err := m.updateNodeObject(cli, node.Name, Labels{}, Annotations{}, ExtendedResources{}, []corev1.Taint{})
if err != nil {
nodeUpdateFailures.Inc()
return fmt.Errorf("failed to prune node %q: %v", node.Name, err)
}
@ -675,6 +676,7 @@ func (m *nfdMaster) SetLabels(c context.Context, r *pb.SetLabelsRequest) (*pb.Se
// Create labels et al
if err := m.refreshNodeFeatures(cli, r.NodeName, annotations, r.GetLabels(), r.GetFeatures()); err != nil {
nodeUpdateFailures.Inc()
return &pb.SetLabelsReply{}, err
}
}

View file

@ -53,6 +53,7 @@ func (u *nodeUpdaterPool) processNodeUpdateRequest(queue workqueue.RateLimitingI
return true
} else {
klog.ErrorS(err, "failed to update node", "nodeName", nodeName)
nodeUpdateFailures.Inc()
}
}
queue.Forget(nodeName)