lib/mergeset: skip createing temporary part objects when merging source inmemory parts

This should reduce CPU usage when adding new entries to inverted index.
This should alos prevent from creating stalled cleaner goroutines for the created temporary parts,
since they were never closed.

This should fix the following issue: https://github.com/VictoriaMetrics/VictoriaMetrics/issues/316 .
This commit is contained in:
Aliaksandr Valialkin 2020-02-13 14:06:51 +02:00
parent 347aaba79d
commit e3b18ca1ab

View file

@ -565,7 +565,7 @@ func (tb *Table) mergeRawItemsBlocks(blocksToMerge []*inmemoryBlock) {
func (tb *Table) mergeInmemoryBlocks(blocksToMerge []*inmemoryBlock) *partWrapper {
// Convert blocksToMerge into inmemoryPart's
pws := make([]*partWrapper, 0, len(blocksToMerge))
mps := make([]*inmemoryPart, 0, len(blocksToMerge))
for _, ib := range blocksToMerge {
if len(ib.items) == 0 {
continue
@ -573,24 +573,25 @@ func (tb *Table) mergeInmemoryBlocks(blocksToMerge []*inmemoryBlock) *partWrappe
mp := getInmemoryPart()
mp.Init(ib)
putInmemoryBlock(ib)
mps = append(mps, mp)
}
if len(mps) == 0 {
return nil
}
if len(mps) == 1 {
// Nothing to merge. Just return a single inmemory part.
mp := mps[0]
p := mp.NewPart()
pw := &partWrapper{
return &partWrapper{
p: p,
mp: mp,
refCount: 1,
}
pws = append(pws, pw)
}
if len(pws) == 0 {
return nil
}
if len(pws) == 1 {
return pws[0]
}
defer func() {
// Return source inmemoryParts to pool.
for _, pw := range pws {
putInmemoryPart(pw.mp)
for _, mp := range mps {
putInmemoryPart(mp)
}
}()
@ -599,10 +600,10 @@ func (tb *Table) mergeInmemoryBlocks(blocksToMerge []*inmemoryBlock) *partWrappe
defer atomic.AddUint64(&tb.activeMerges, ^uint64(0))
// Prepare blockStreamReaders for source parts.
bsrs := make([]*blockStreamReader, 0, len(pws))
for _, pw := range pws {
bsrs := make([]*blockStreamReader, 0, len(mps))
for _, mp := range mps {
bsr := getBlockStreamReader()
bsr.InitFromInmemoryPart(pw.mp)
bsr.InitFromInmemoryPart(mp)
bsrs = append(bsrs, bsr)
}