prometheus/tsdb/ooo_head_read.go
Arve Knudsen 06d54fcc6c
[PERF] TSDB: Optimize inverse matching (#14144)
Simple follow-up to #13620. Modify `tsdb.PostingsForMatchers` to use the optimized tsdb.IndexReader.PostingsForLabelMatching method also for inverse matching.

Introduce method `PostingsForAllLabelValues`, to avoid changing the existing method.

The performance is much improved for a subset of the cases; there are up to
~60% CPU gains and ~12.5% reduction in memory usage. 

Remove `TestReader_InversePostingsForMatcherHonorsContextCancel` since
`inversePostingsForMatcher` only passes `ctx` to `IndexReader` implementations now.

Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
2024-11-19 15:49:01 +00:00

618 lines
20 KiB
Go

// Copyright 2022 The Prometheus Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package tsdb
import (
"context"
"errors"
"fmt"
"math"
"slices"
"github.com/oklog/ulid"
"github.com/prometheus/prometheus/model/labels"
"github.com/prometheus/prometheus/storage"
"github.com/prometheus/prometheus/tsdb/chunkenc"
"github.com/prometheus/prometheus/tsdb/chunks"
"github.com/prometheus/prometheus/tsdb/index"
"github.com/prometheus/prometheus/tsdb/tombstones"
"github.com/prometheus/prometheus/util/annotations"
)
var _ IndexReader = &HeadAndOOOIndexReader{}
type HeadAndOOOIndexReader struct {
*headIndexReader // A reference to the headIndexReader so we can reuse as many interface implementation as possible.
inoMint int64
lastGarbageCollectedMmapRef chunks.ChunkDiskMapperRef
}
var _ chunkenc.Iterable = &mergedOOOChunks{}
// mergedOOOChunks holds the list of iterables for overlapping chunks.
type mergedOOOChunks struct {
chunkIterables []chunkenc.Iterable
}
func (o mergedOOOChunks) Iterator(iterator chunkenc.Iterator) chunkenc.Iterator {
return storage.ChainSampleIteratorFromIterables(iterator, o.chunkIterables)
}
func NewHeadAndOOOIndexReader(head *Head, inoMint, mint, maxt int64, lastGarbageCollectedMmapRef chunks.ChunkDiskMapperRef) *HeadAndOOOIndexReader {
hr := &headIndexReader{
head: head,
mint: mint,
maxt: maxt,
}
return &HeadAndOOOIndexReader{hr, inoMint, lastGarbageCollectedMmapRef}
}
func (oh *HeadAndOOOIndexReader) Series(ref storage.SeriesRef, builder *labels.ScratchBuilder, chks *[]chunks.Meta) error {
s := oh.head.series.getByID(chunks.HeadSeriesRef(ref))
if s == nil {
oh.head.metrics.seriesNotFound.Inc()
return storage.ErrNotFound
}
builder.Assign(s.labels())
if chks == nil {
return nil
}
s.Lock()
defer s.Unlock()
*chks = (*chks)[:0]
if s.ooo != nil {
return getOOOSeriesChunks(s, oh.mint, oh.maxt, oh.lastGarbageCollectedMmapRef, 0, true, oh.inoMint, chks)
}
*chks = appendSeriesChunks(s, oh.inoMint, oh.maxt, *chks)
return nil
}
// lastGarbageCollectedMmapRef gives the last mmap chunk that may be being garbage collected and so
// any chunk at or before this ref will not be considered. 0 disables this check.
//
// maxMmapRef tells upto what max m-map chunk that we can consider. If it is non-0, then
// the oooHeadChunk will not be considered.
func getOOOSeriesChunks(s *memSeries, mint, maxt int64, lastGarbageCollectedMmapRef, maxMmapRef chunks.ChunkDiskMapperRef, includeInOrder bool, inoMint int64, chks *[]chunks.Meta) error {
tmpChks := make([]chunks.Meta, 0, len(s.ooo.oooMmappedChunks))
addChunk := func(minT, maxT int64, ref chunks.ChunkRef, chunk chunkenc.Chunk) {
tmpChks = append(tmpChks, chunks.Meta{
MinTime: minT,
MaxTime: maxT,
Ref: ref,
Chunk: chunk,
})
}
// Collect all chunks that overlap the query range.
if s.ooo.oooHeadChunk != nil {
c := s.ooo.oooHeadChunk
if c.OverlapsClosedInterval(mint, maxt) && maxMmapRef == 0 {
ref := chunks.ChunkRef(chunks.NewHeadChunkRef(s.ref, s.oooHeadChunkID(len(s.ooo.oooMmappedChunks))))
if len(c.chunk.samples) > 0 { // Empty samples happens in tests, at least.
chks, err := s.ooo.oooHeadChunk.chunk.ToEncodedChunks(c.minTime, c.maxTime)
if err != nil {
handleChunkWriteError(err)
return nil
}
for _, chk := range chks {
addChunk(chk.minTime, chk.maxTime, ref, chk.chunk)
}
} else {
var emptyChunk chunkenc.Chunk
addChunk(c.minTime, c.maxTime, ref, emptyChunk)
}
}
}
for i := len(s.ooo.oooMmappedChunks) - 1; i >= 0; i-- {
c := s.ooo.oooMmappedChunks[i]
if c.OverlapsClosedInterval(mint, maxt) && (maxMmapRef == 0 || maxMmapRef.GreaterThanOrEqualTo(c.ref)) && (lastGarbageCollectedMmapRef == 0 || c.ref.GreaterThan(lastGarbageCollectedMmapRef)) {
ref := chunks.ChunkRef(chunks.NewHeadChunkRef(s.ref, s.oooHeadChunkID(i)))
addChunk(c.minTime, c.maxTime, ref, nil)
}
}
if includeInOrder {
tmpChks = appendSeriesChunks(s, inoMint, maxt, tmpChks)
}
// There is nothing to do if we did not collect any chunk.
if len(tmpChks) == 0 {
return nil
}
// Next we want to sort all the collected chunks by min time so we can find
// those that overlap.
slices.SortFunc(tmpChks, lessByMinTimeAndMinRef)
// Next we want to iterate the sorted collected chunks and return composites for chunks that overlap with others.
// Example chunks of a series: 5:(100, 200) 6:(500, 600) 7:(150, 250) 8:(550, 650)
// In the example 5 overlaps with 7 and 6 overlaps with 8 so we will return
// [5,7], [6,8].
toBeMerged := tmpChks[0]
for _, c := range tmpChks[1:] {
if c.MinTime > toBeMerged.MaxTime {
// This chunk doesn't overlap. Send current toBeMerged to output and start a new one.
*chks = append(*chks, toBeMerged)
toBeMerged = c
} else {
// Merge this chunk with existing toBeMerged.
if mm, ok := toBeMerged.Chunk.(*multiMeta); ok {
mm.metas = append(mm.metas, c)
} else {
toBeMerged.Chunk = &multiMeta{metas: []chunks.Meta{toBeMerged, c}}
}
if toBeMerged.MaxTime < c.MaxTime {
toBeMerged.MaxTime = c.MaxTime
}
}
}
*chks = append(*chks, toBeMerged)
return nil
}
// Fake Chunk object to pass a set of Metas inside Meta.Chunk.
type multiMeta struct {
chunkenc.Chunk // We don't expect any of the methods to be called.
metas []chunks.Meta
}
// LabelValues needs to be overridden from the headIndexReader implementation
// so we can return labels within either in-order range or ooo range.
func (oh *HeadAndOOOIndexReader) LabelValues(ctx context.Context, name string, matchers ...*labels.Matcher) ([]string, error) {
if oh.maxt < oh.head.MinTime() && oh.maxt < oh.head.MinOOOTime() || oh.mint > oh.head.MaxTime() && oh.mint > oh.head.MaxOOOTime() {
return []string{}, nil
}
if len(matchers) == 0 {
return oh.head.postings.LabelValues(ctx, name), nil
}
return labelValuesWithMatchers(ctx, oh, name, matchers...)
}
func lessByMinTimeAndMinRef(a, b chunks.Meta) int {
switch {
case a.MinTime < b.MinTime:
return -1
case a.MinTime > b.MinTime:
return 1
}
switch {
case a.Ref < b.Ref:
return -1
case a.Ref > b.Ref:
return 1
default:
return 0
}
}
type HeadAndOOOChunkReader struct {
head *Head
mint, maxt int64
cr *headChunkReader // If nil, only read OOO chunks.
maxMmapRef chunks.ChunkDiskMapperRef
oooIsoState *oooIsolationState
}
func NewHeadAndOOOChunkReader(head *Head, mint, maxt int64, cr *headChunkReader, oooIsoState *oooIsolationState, maxMmapRef chunks.ChunkDiskMapperRef) *HeadAndOOOChunkReader {
return &HeadAndOOOChunkReader{
head: head,
mint: mint,
maxt: maxt,
cr: cr,
maxMmapRef: maxMmapRef,
oooIsoState: oooIsoState,
}
}
func (cr *HeadAndOOOChunkReader) ChunkOrIterable(meta chunks.Meta) (chunkenc.Chunk, chunkenc.Iterable, error) {
c, it, _, err := cr.chunkOrIterable(meta, false)
return c, it, err
}
// ChunkOrIterableWithCopy implements ChunkReaderWithCopy. The special Copy
// behaviour is only implemented for the in-order head chunk.
func (cr *HeadAndOOOChunkReader) ChunkOrIterableWithCopy(meta chunks.Meta) (chunkenc.Chunk, chunkenc.Iterable, int64, error) {
return cr.chunkOrIterable(meta, true)
}
func (cr *HeadAndOOOChunkReader) chunkOrIterable(meta chunks.Meta, copyLastChunk bool) (chunkenc.Chunk, chunkenc.Iterable, int64, error) {
sid, cid, isOOO := unpackHeadChunkRef(meta.Ref)
s := cr.head.series.getByID(sid)
// This means that the series has been garbage collected.
if s == nil {
return nil, nil, 0, storage.ErrNotFound
}
var isoState *isolationState
if cr.cr != nil {
isoState = cr.cr.isoState
}
s.Lock()
defer s.Unlock()
if meta.Chunk == nil {
c, maxt, err := cr.head.chunkFromSeries(s, cid, isOOO, meta.MinTime, meta.MaxTime, isoState, copyLastChunk)
return c, nil, maxt, err
}
mm, ok := meta.Chunk.(*multiMeta)
if !ok { // Complete chunk was supplied.
return meta.Chunk, nil, meta.MaxTime, nil
}
// We have a composite meta: construct a composite iterable.
mc := &mergedOOOChunks{}
for _, m := range mm.metas {
switch {
case m.Chunk != nil:
mc.chunkIterables = append(mc.chunkIterables, m.Chunk)
default:
_, cid, isOOO := unpackHeadChunkRef(m.Ref)
iterable, _, err := cr.head.chunkFromSeries(s, cid, isOOO, m.MinTime, m.MaxTime, isoState, copyLastChunk)
if err != nil {
return nil, nil, 0, fmt.Errorf("invalid head chunk: %w", err)
}
mc.chunkIterables = append(mc.chunkIterables, iterable)
}
}
return nil, mc, meta.MaxTime, nil
}
func (cr *HeadAndOOOChunkReader) Close() error {
if cr.cr != nil && cr.cr.isoState != nil {
cr.cr.isoState.Close()
}
if cr.oooIsoState != nil {
cr.oooIsoState.Close()
}
return nil
}
type OOOCompactionHead struct {
head *Head
lastMmapRef chunks.ChunkDiskMapperRef
lastWBLFile int
postings []storage.SeriesRef
chunkRange int64
mint, maxt int64 // Among all the compactable chunks.
}
// NewOOOCompactionHead does the following:
// 1. M-maps all the in-memory ooo chunks.
// 2. Compute the expected block ranges while iterating through all ooo series and store it.
// 3. Store the list of postings having ooo series.
// 4. Cuts a new WBL file for the OOO WBL.
// All the above together have a bit of CPU and memory overhead, and can have a bit of impact
// on the sample append latency. So call NewOOOCompactionHead only right before compaction.
func NewOOOCompactionHead(ctx context.Context, head *Head) (*OOOCompactionHead, error) {
ch := &OOOCompactionHead{
head: head,
chunkRange: head.chunkRange.Load(),
mint: math.MaxInt64,
maxt: math.MinInt64,
}
if head.wbl != nil {
lastWBLFile, err := head.wbl.NextSegmentSync()
if err != nil {
return nil, err
}
ch.lastWBLFile = lastWBLFile
}
hr := headIndexReader{head: head, mint: ch.mint, maxt: ch.maxt}
n, v := index.AllPostingsKey()
// TODO: filter to series with OOO samples, before sorting.
p, err := hr.Postings(ctx, n, v)
if err != nil {
return nil, err
}
p = hr.SortedPostings(p)
var lastSeq, lastOff int
for p.Next() {
seriesRef := p.At()
ms := head.series.getByID(chunks.HeadSeriesRef(seriesRef))
if ms == nil {
continue
}
// M-map the in-memory chunk and keep track of the last one.
// Also build the block ranges -> series map.
// TODO: consider having a lock specifically for ooo data.
ms.Lock()
if ms.ooo == nil {
ms.Unlock()
continue
}
var lastMmapRef chunks.ChunkDiskMapperRef
mmapRefs := ms.mmapCurrentOOOHeadChunk(head.chunkDiskMapper, head.logger)
if len(mmapRefs) == 0 && len(ms.ooo.oooMmappedChunks) > 0 {
// Nothing was m-mapped. So take the mmapRef from the existing slice if it exists.
mmapRefs = []chunks.ChunkDiskMapperRef{ms.ooo.oooMmappedChunks[len(ms.ooo.oooMmappedChunks)-1].ref}
}
if len(mmapRefs) == 0 {
lastMmapRef = 0
} else {
lastMmapRef = mmapRefs[len(mmapRefs)-1]
}
seq, off := lastMmapRef.Unpack()
if seq > lastSeq || (seq == lastSeq && off > lastOff) {
ch.lastMmapRef, lastSeq, lastOff = lastMmapRef, seq, off
}
if len(ms.ooo.oooMmappedChunks) > 0 {
ch.postings = append(ch.postings, seriesRef)
for _, c := range ms.ooo.oooMmappedChunks {
if c.minTime < ch.mint {
ch.mint = c.minTime
}
if c.maxTime > ch.maxt {
ch.maxt = c.maxTime
}
}
}
ms.Unlock()
}
return ch, nil
}
func (ch *OOOCompactionHead) Index() (IndexReader, error) {
return NewOOOCompactionHeadIndexReader(ch), nil
}
func (ch *OOOCompactionHead) Chunks() (ChunkReader, error) {
return NewHeadAndOOOChunkReader(ch.head, ch.mint, ch.maxt, nil, nil, ch.lastMmapRef), nil
}
func (ch *OOOCompactionHead) Tombstones() (tombstones.Reader, error) {
return tombstones.NewMemTombstones(), nil
}
var oooCompactionHeadULID = ulid.MustParse("0000000000XX000COMPACTHEAD")
func (ch *OOOCompactionHead) Meta() BlockMeta {
return BlockMeta{
MinTime: ch.mint,
MaxTime: ch.maxt,
ULID: oooCompactionHeadULID,
Stats: BlockStats{
NumSeries: uint64(len(ch.postings)),
},
}
}
// CloneForTimeRange clones the OOOCompactionHead such that the IndexReader and ChunkReader
// obtained from this only looks at the m-map chunks within the given time ranges while not looking
// beyond the ch.lastMmapRef.
// Only the method of BlockReader interface are valid for the cloned OOOCompactionHead.
func (ch *OOOCompactionHead) CloneForTimeRange(mint, maxt int64) *OOOCompactionHead {
return &OOOCompactionHead{
head: ch.head,
lastMmapRef: ch.lastMmapRef,
postings: ch.postings,
chunkRange: ch.chunkRange,
mint: mint,
maxt: maxt,
}
}
func (ch *OOOCompactionHead) Size() int64 { return 0 }
func (ch *OOOCompactionHead) MinTime() int64 { return ch.mint }
func (ch *OOOCompactionHead) MaxTime() int64 { return ch.maxt }
func (ch *OOOCompactionHead) ChunkRange() int64 { return ch.chunkRange }
func (ch *OOOCompactionHead) LastMmapRef() chunks.ChunkDiskMapperRef { return ch.lastMmapRef }
func (ch *OOOCompactionHead) LastWBLFile() int { return ch.lastWBLFile }
type OOOCompactionHeadIndexReader struct {
ch *OOOCompactionHead
}
func NewOOOCompactionHeadIndexReader(ch *OOOCompactionHead) IndexReader {
return &OOOCompactionHeadIndexReader{ch: ch}
}
func (ir *OOOCompactionHeadIndexReader) Symbols() index.StringIter {
hr := headIndexReader{head: ir.ch.head, mint: ir.ch.mint, maxt: ir.ch.maxt}
return hr.Symbols()
}
func (ir *OOOCompactionHeadIndexReader) Postings(_ context.Context, name string, values ...string) (index.Postings, error) {
n, v := index.AllPostingsKey()
if name != n || len(values) != 1 || values[0] != v {
return nil, errors.New("only AllPostingsKey is supported")
}
return index.NewListPostings(ir.ch.postings), nil
}
func (ir *OOOCompactionHeadIndexReader) PostingsForLabelMatching(context.Context, string, func(string) bool) index.Postings {
return index.ErrPostings(errors.New("not supported"))
}
func (ir *OOOCompactionHeadIndexReader) PostingsForAllLabelValues(context.Context, string) index.Postings {
return index.ErrPostings(errors.New("not supported"))
}
func (ir *OOOCompactionHeadIndexReader) SortedPostings(p index.Postings) index.Postings {
// This will already be sorted from the Postings() call above.
return p
}
func (ir *OOOCompactionHeadIndexReader) ShardedPostings(p index.Postings, shardIndex, shardCount uint64) index.Postings {
hr := headIndexReader{head: ir.ch.head, mint: ir.ch.mint, maxt: ir.ch.maxt}
return hr.ShardedPostings(p, shardIndex, shardCount)
}
func (ir *OOOCompactionHeadIndexReader) Series(ref storage.SeriesRef, builder *labels.ScratchBuilder, chks *[]chunks.Meta) error {
s := ir.ch.head.series.getByID(chunks.HeadSeriesRef(ref))
if s == nil {
ir.ch.head.metrics.seriesNotFound.Inc()
return storage.ErrNotFound
}
builder.Assign(s.labels())
s.Lock()
defer s.Unlock()
*chks = (*chks)[:0]
if s.ooo == nil {
return nil
}
return getOOOSeriesChunks(s, ir.ch.mint, ir.ch.maxt, 0, ir.ch.lastMmapRef, false, 0, chks)
}
func (ir *OOOCompactionHeadIndexReader) SortedLabelValues(_ context.Context, name string, matchers ...*labels.Matcher) ([]string, error) {
return nil, errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) LabelValues(_ context.Context, name string, matchers ...*labels.Matcher) ([]string, error) {
return nil, errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) PostingsForMatchers(_ context.Context, concurrent bool, ms ...*labels.Matcher) (index.Postings, error) {
return nil, errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) LabelNames(context.Context, ...*labels.Matcher) ([]string, error) {
return nil, errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) LabelValueFor(context.Context, storage.SeriesRef, string) (string, error) {
return "", errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) LabelNamesFor(ctx context.Context, postings index.Postings) ([]string, error) {
return nil, errors.New("not implemented")
}
func (ir *OOOCompactionHeadIndexReader) Close() error {
return nil
}
// HeadAndOOOQuerier queries both the head and the out-of-order head.
type HeadAndOOOQuerier struct {
mint, maxt int64
head *Head
index IndexReader
chunkr ChunkReader
querier storage.Querier // Used for LabelNames, LabelValues, but may be nil if head was truncated in the mean time, in which case we ignore it and not close it in the end.
}
func NewHeadAndOOOQuerier(inoMint, mint, maxt int64, head *Head, oooIsoState *oooIsolationState, querier storage.Querier) storage.Querier {
cr := &headChunkReader{
head: head,
mint: mint,
maxt: maxt,
isoState: head.iso.State(mint, maxt),
}
return &HeadAndOOOQuerier{
mint: mint,
maxt: maxt,
head: head,
index: NewHeadAndOOOIndexReader(head, inoMint, mint, maxt, oooIsoState.minRef),
chunkr: NewHeadAndOOOChunkReader(head, mint, maxt, cr, oooIsoState, 0),
querier: querier,
}
}
func (q *HeadAndOOOQuerier) LabelValues(ctx context.Context, name string, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
if q.querier == nil {
return nil, nil, nil
}
return q.querier.LabelValues(ctx, name, hints, matchers...)
}
func (q *HeadAndOOOQuerier) LabelNames(ctx context.Context, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
if q.querier == nil {
return nil, nil, nil
}
return q.querier.LabelNames(ctx, hints, matchers...)
}
func (q *HeadAndOOOQuerier) Close() error {
q.chunkr.Close()
if q.querier == nil {
return nil
}
return q.querier.Close()
}
func (q *HeadAndOOOQuerier) Select(ctx context.Context, sortSeries bool, hints *storage.SelectHints, matchers ...*labels.Matcher) storage.SeriesSet {
return selectSeriesSet(ctx, sortSeries, hints, matchers, q.index, q.chunkr, q.head.tombstones, q.mint, q.maxt)
}
// HeadAndOOOChunkQuerier queries both the head and the out-of-order head.
type HeadAndOOOChunkQuerier struct {
mint, maxt int64
head *Head
index IndexReader
chunkr ChunkReader
querier storage.ChunkQuerier
}
func NewHeadAndOOOChunkQuerier(inoMint, mint, maxt int64, head *Head, oooIsoState *oooIsolationState, querier storage.ChunkQuerier) storage.ChunkQuerier {
cr := &headChunkReader{
head: head,
mint: mint,
maxt: maxt,
isoState: head.iso.State(mint, maxt),
}
return &HeadAndOOOChunkQuerier{
mint: mint,
maxt: maxt,
head: head,
index: NewHeadAndOOOIndexReader(head, inoMint, mint, maxt, oooIsoState.minRef),
chunkr: NewHeadAndOOOChunkReader(head, mint, maxt, cr, oooIsoState, 0),
querier: querier,
}
}
func (q *HeadAndOOOChunkQuerier) LabelValues(ctx context.Context, name string, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
if q.querier == nil {
return nil, nil, nil
}
return q.querier.LabelValues(ctx, name, hints, matchers...)
}
func (q *HeadAndOOOChunkQuerier) LabelNames(ctx context.Context, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
if q.querier == nil {
return nil, nil, nil
}
return q.querier.LabelNames(ctx, hints, matchers...)
}
func (q *HeadAndOOOChunkQuerier) Close() error {
q.chunkr.Close()
if q.querier == nil {
return nil
}
return q.querier.Close()
}
func (q *HeadAndOOOChunkQuerier) Select(ctx context.Context, sortSeries bool, hints *storage.SelectHints, matchers ...*labels.Matcher) storage.ChunkSeriesSet {
return selectChunkSeriesSet(ctx, sortSeries, hints, matchers, rangeHeadULID, q.index, q.chunkr, q.head.tombstones, q.mint, q.maxt)
}