Skip to content

statistics: delete cluster status metrics on store deletion #9945

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
ti-chi-bot merged 3 commits into from
Nov 21, 2025
Merged

statistics: delete cluster status metrics on store deletion #9945

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
cd7a575
statistics: add engine type handling and delete cluster status metric…
okJiang Nov 19, 2025
612d7e8
add comment
okJiang Nov 19, 2025
05d7b94
fix lint
okJiang Nov 19, 2025
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
8 changes: 8 additions & 0 deletions pkg/core/store.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -244,6 +244,14 @@ func (s *StoreInfo) IsTiFlashCompute() bool {
return IsStoreContainLabel(s.GetMeta(), EngineKey, EngineTiFlashCompute)
}

// Engine returns the engine type of the store.
func (s *StoreInfo) Engine() string {
if s.IsTiKV() {
return EngineTiKV
}
return EngineTiFlash
}

// IsUp returns true if store is serving or preparing.
func (s *StoreInfo) IsUp() bool {
return s.IsServing() || s.IsPreparing()
Expand Down
1 change: 1 addition & 0 deletions pkg/mcs/router/server/meta/watcher.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,6 +95,7 @@ func (w *Watcher) initializeStoreWatcher() error {
}
origin := w.basicCluster.GetStore(storeID)
if origin != nil {
statistics.DeleteClusterStatusMetrics(origin)
w.basicCluster.DeleteStore(origin)
log.Info("delete store meta", zap.Uint64("store-id", storeID))
}
Expand Down
1 change: 1 addition & 0 deletions pkg/mcs/scheduling/server/meta/watcher.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,6 +95,7 @@ func (w *Watcher) initializeStoreWatcher() error {
}
origin := w.basicCluster.GetStore(storeID)
if origin != nil {
statistics.DeleteClusterStatusMetrics(origin)
w.basicCluster.DeleteStore(origin)
log.Info("delete store meta", zap.Uint64("store-id", storeID))
}
Expand Down
16 changes: 15 additions & 1 deletion pkg/statistics/metrics.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,13 @@

package statistics

import "github.com/prometheus/client_golang/prometheus"
import (
"strconv"

"github.com/prometheus/client_golang/prometheus"

"github.com/tikv/pd/pkg/core"
)

var (
hotCacheStatusGauge = prometheus.NewGaugeVec(
Expand Down Expand Up @@ -118,3 +124,11 @@ func init() {
prometheus.MustRegister(regionAbnormalPeerDuration)
prometheus.MustRegister(hotPeerSummary)
}

// DeleteClusterStatusMetrics deletes the cluster status metrics of a store.
func DeleteClusterStatusMetrics(store *core.StoreInfo) {
engine := store.Engine()
for _, status := range storeStatuses {
clusterStatusGauge.DeleteLabelValues(status, engine, strconv.FormatUint(store.GetID(), 10))
}
}
22 changes: 16 additions & 6 deletions pkg/statistics/store_collection.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,6 +52,21 @@ const (
clusterStatusStorageCapacity = "storage_capacity"
)

var storeStatuses = []string{
clusterStatusStoreUpCount,
clusterStatusStoreDisconnectedCount,
clusterStatusStoreSlowCount,
clusterStatusStoreDownCount,
clusterStatusStoreUnhealthCount,
clusterStatusStoreOfflineCount,
clusterStatusStoreTombstoneCount,
clusterStatusStoreLowSpaceCount,
clusterStatusStorePreparingCount,
clusterStatusStoreServingCount,
clusterStatusStoreRemovingCount,
clusterStatusStoreRemovedCount,
}

type storeStatistics struct {
opt config.ConfProvider
LabelCounter map[string][]uint64
Expand Down Expand Up @@ -129,12 +144,7 @@ func (s *storeStatistics) observe(store *core.StoreInfo) {
}
storeAddress := store.GetAddress()
id := strconv.FormatUint(store.GetID(), 10)
var engine string
if store.IsTiKV() {
engine = core.EngineTiKV
} else {
engine = core.EngineTiFlash
}
engine := store.Engine()
storeStatusStats := s.observeStoreStatus(store)
for statusType, value := range storeStatusStats {
clusterStatusGauge.WithLabelValues(statusType, engine, id).Set(value)
Expand Down
1 change: 1 addition & 0 deletions server/cluster/cluster.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2008,6 +2008,7 @@ func (c *RaftCluster) deleteStore(store *core.StoreInfo) error {
return err
}
}
statistics.DeleteClusterStatusMetrics(store)
c.DeleteStore(store)
return nil
}
Expand Down