Skip to content

MB-72082: improve stats tracking on zap layer #2392

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

Open
Thejas-bhat wants to merge 4 commits into
base: master
Choose a base branch
from
+21 −10
Open

MB-72082: improve stats tracking on zap layer #2392

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
35a3e04
stats
Thejas-bhat Jun 9, 2026
2426b82
add more training stats
Thejas-bhat Jul 7, 2026
ffa9539
fix the zap stats to be per bleve index
Thejas-bhat Aug 19, 2026
161ed7c
go.mod update
Thejas-bhat Aug 21, 2026
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
4 changes: 2 additions & 2 deletions go.mod
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ require (
github.com/blevesearch/go-porterstemmer v1.0.3
github.com/blevesearch/goleveldb v1.0.1
github.com/blevesearch/gtreap v0.1.1
github.com/blevesearch/scorch_segment_api/v2 v2.4.9-0.20260729090843-4313bda09bee
github.com/blevesearch/scorch_segment_api/v2 v2.4.9
github.com/blevesearch/segment v0.9.1
github.com/blevesearch/snowball v0.6.1
github.com/blevesearch/snowballstem v0.9.0
Expand All @@ -25,7 +25,7 @@ require (
github.com/blevesearch/zapx/v14 v14.4.3
github.com/blevesearch/zapx/v15 v15.4.3
github.com/blevesearch/zapx/v16 v16.3.4
github.com/blevesearch/zapx/v17 v17.2.2-0.20260806075952-6594f3d6bd0d
github.com/blevesearch/zapx/v17 v17.2.2
github.com/couchbase/moss v0.2.0
github.com/spf13/cobra v1.10.2
go.etcd.io/bbolt v1.4.0
Expand Down
8 changes: 4 additions & 4 deletions go.sum
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,8 +19,8 @@ github.com/blevesearch/gtreap v0.1.1/go.mod h1:QaQyDRAT51sotthUWAH4Sj08awFSSWzgY
github.com/blevesearch/mmap-go v1.0.2/go.mod h1:ol2qBqYaOUsGdm7aRMRrYGgPvnwLe6Y+7LMvAB5IbSA=
github.com/blevesearch/mmap-go v1.2.0 h1:l33nNKPFcBjJUMwem6sAYJPUzhUCABoK9FxZDGiFNBI=
github.com/blevesearch/mmap-go v1.2.0/go.mod h1:Vd6+20GBhEdwJnU1Xohgt88XCD/CTWcqbCNxkZpyBo0=
github.com/blevesearch/scorch_segment_api/v2 v2.4.9-0.20260729090843-4313bda09bee h1:RmX8uCxp4RDbsW0PRFsFED5f7X5xUvz6/9cYb9HrNTU=
github.com/blevesearch/scorch_segment_api/v2 v2.4.9-0.20260729090843-4313bda09bee/go.mod h1:WUUkAocbkDlNK/kgAE13NvS9oxe+u618mYZ8sOvcCc4=
github.com/blevesearch/scorch_segment_api/v2 v2.4.9 h1:q1shQvpwVGp0IISmX08Ead5XFOPNmZrB761Snpru3vk=
github.com/blevesearch/scorch_segment_api/v2 v2.4.9/go.mod h1:WUUkAocbkDlNK/kgAE13NvS9oxe+u618mYZ8sOvcCc4=
github.com/blevesearch/segment v0.9.1 h1:+dThDy+Lvgj5JMxhmOVlgFfkUtZV2kw49xax4+jTfSU=
github.com/blevesearch/segment v0.9.1/go.mod h1:zN21iLm7+GnBHWTao9I+Au/7MBiL8pPFtJBJTsk6kQw=
github.com/blevesearch/snowball v0.6.1 h1:cDYjn/NCH+wwt2UdehaLpr2e4BwLIjN4V/TdLsL+B5A=
Expand All @@ -45,8 +45,8 @@ github.com/blevesearch/zapx/v15 v15.4.3 h1:iJiMJOHrz216jyO6lS0m9RTCEkprUnzvqAI2l
github.com/blevesearch/zapx/v15 v15.4.3/go.mod h1:1pssev/59FsuWcgSnTa0OeEpOzmhtmr/0/11H0Z8+Nw=
github.com/blevesearch/zapx/v16 v16.3.4 h1:hDAqA8qusZTNbPEL7//w5P65UZ2de6yhSeUaTbp0Po0=
github.com/blevesearch/zapx/v16 v16.3.4/go.mod h1:zqkPPqs9GS9FzVWzCO3Wf1X044yWAV17+4zb+FTiEHg=
github.com/blevesearch/zapx/v17 v17.2.2-0.20260806075952-6594f3d6bd0d h1:NsgFVBOgq3lg7ClW4e7iJc5OsW7MFlsSIGxiGW7++Fw=
github.com/blevesearch/zapx/v17 v17.2.2-0.20260806075952-6594f3d6bd0d/go.mod h1:hPt57M7CxaMgVQF1lYak7rRaO7Ge5j/sdYzrs8t8Htk=
github.com/blevesearch/zapx/v17 v17.2.2 h1:v3as9jp2CFsrPKpBotYWn/Evp/EjElH4xpNvOhgsNTU=
github.com/blevesearch/zapx/v17 v17.2.2/go.mod h1:r/ddAUuomMMcx12ZUBoDxf/9Fa4aqvFATf7AJzWp/Ac=
github.com/couchbase/ghistogram v0.1.0 h1:b95QcQTCzjTUocDXp/uMgSNQi8oj1tGwnJ4bODWZnps=
github.com/couchbase/ghistogram v0.1.0/go.mod h1:s1Jhy76zqfEecpNWJfWUiKZookAFaiGOEoyzgHt9i7k=
github.com/couchbase/moss v0.2.0 h1:VCYrMzFwEryyhRSeI+/b3tRBSeTpi/8gn5Kf6dxqn+o=
Expand Down
7 changes: 7 additions & 0 deletions index/scorch/scorch.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@ var ErrClosed = fmt.Errorf("scorch closed")
type Scorch struct {
nextSegmentID uint64
stats Stats
zapStats segment.Stats
iStats internalStats

readOnly bool
Expand Down Expand Up @@ -213,6 +214,9 @@ func NewScorch(storeName string,
rv.segmentConfig = segConfig
}

// NOTE: always register the stats handler in the config map to track zap layer stats
rv.segmentConfig[segment.StatsKey] = &rv.zapStats

typ, ok := config["spatialPlugin"].(string)
if ok {
if err := rv.loadSpatialAnalyzerPlugin(typ); err != nil {
Expand Down Expand Up @@ -853,6 +857,9 @@ func (s *Scorch) StatsMap() map[string]interface{} {
m["field:"+fieldName+":"+statName] = val
}
}

// zap layer stats that are continously updated throughout the index lifecycle
m[segment.StatsKey] = s.zapStats.StatsMap()
return m
}

Expand Down
6 changes: 4 additions & 2 deletions index/scorch/stats.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,8 +144,10 @@ type Stats struct {

TotMemorySegmentsAtRoot uint64

TotTrainedSamples uint64
TotTrainTime uint64
TotTrainedSamples uint64
TotTrainTime uint64
TotTrainFireIndexEventTime uint64
TotTrainFireIndexEvents uint64
}

// atomically populates the returned map
Expand Down
6 changes: 4 additions & 2 deletions index/scorch/train_vector.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,7 +69,7 @@ func initTrainer(s *Scorch, config map[string]interface{}) *vectorTrainer {
if ok && feature {
trainer := vectorTrainer{
parent: s,
config: maps.Clone(s.config),
config: maps.Clone(s.segmentConfig),
trainCh: make(chan *trainRequest, 1),
doneCh: make(chan struct{}),
}
Expand Down Expand Up @@ -340,7 +340,10 @@ func (t *vectorTrainer) loadTrainedData(bucket *util.BoltBucketImpl) error {

func (t *vectorTrainer) train(batch *index.Batch) error {
// regulate the Train function
start := time.Now()
t.parent.FireIndexEvent()
atomic.AddUint64(&t.parent.stats.TotTrainFireIndexEventTime, uint64(time.Since(start)))
atomic.AddUint64(&t.parent.stats.TotTrainFireIndexEvents, 1)
if t.trainingComplete.Load() {
return fmt.Errorf("training is already complete, cannot accept more training data")
}
Expand Down Expand Up @@ -393,7 +396,6 @@ func (t *vectorTrainer) train(batch *index.Batch) error {
// is complete, the template will be used for other operations down the line
// like merge and search.
//
// note: this might index text data too, how to handle this? s.segmentConfig?
// todo: updates/deletes -> data drift detection
if len(trainData) > 0 {
trainReq.sample, _, err = t.parent.segPlugin.NewUsing(trainData, config)
Expand Down
Loading