milvus/internal/indexcoord/garbage_collector.go
SimFG a55f739608
Separate public proto files (#19782)
Signed-off-by: SimFG <bang.fu@zilliz.com>

Signed-off-by: SimFG <bang.fu@zilliz.com>
2022-10-16 20:49:27 +08:00

280 lines
9.9 KiB
Go

// Licensed to the LF AI & Data foundation under one
// or more contributor license agreements. See the NOTICE file
// distributed with this work for additional information
// regarding copyright ownership. The ASF licenses this file
// to you 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 indexcoord
import (
"context"
"path"
"sync"
"time"
"go.uber.org/zap"
"github.com/milvus-io/milvus-proto/go-api/commonpb"
"github.com/milvus-io/milvus/internal/common"
"github.com/milvus-io/milvus/internal/log"
"github.com/milvus-io/milvus/internal/metastore/model"
"github.com/milvus-io/milvus/internal/proto/datapb"
"github.com/milvus-io/milvus/internal/storage"
)
type garbageCollector struct {
ctx context.Context
cancel context.CancelFunc
wg sync.WaitGroup
gcFileDuration time.Duration
gcMetaDuration time.Duration
metaTable *metaTable
chunkManager storage.ChunkManager
indexCoordClient *IndexCoord
}
func newGarbageCollector(ctx context.Context, meta *metaTable, chunkManager storage.ChunkManager, ic *IndexCoord) *garbageCollector {
ctx, cancel := context.WithCancel(ctx)
return &garbageCollector{
ctx: ctx,
cancel: cancel,
gcFileDuration: Params.IndexCoordCfg.GCInterval,
gcMetaDuration: time.Minute,
metaTable: meta,
chunkManager: chunkManager,
indexCoordClient: ic,
}
}
func (gc *garbageCollector) Start() {
gc.wg.Add(1)
go gc.recycleUnusedIndexes()
gc.wg.Add(1)
go gc.recycleUnusedSegIndexes()
gc.wg.Add(1)
go gc.recycleUnusedIndexFiles()
}
func (gc *garbageCollector) Stop() {
gc.cancel()
gc.wg.Wait()
}
func (gc *garbageCollector) recycleUnusedIndexes() {
defer gc.wg.Done()
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector recycleUnusedIndexes start")
ticker := time.NewTicker(gc.gcMetaDuration)
defer ticker.Stop()
for {
select {
case <-gc.ctx.Done():
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector recycleUnusedMetaLoop context has done")
return
case <-ticker.C:
deletedIndexes := gc.metaTable.GetDeletedIndexes()
for _, index := range deletedIndexes {
buildIDs := gc.metaTable.GetBuildIDsFromIndexID(index.IndexID)
if len(buildIDs) == 0 {
if err := gc.metaTable.RemoveIndex(index.CollectionID, index.IndexID); err != nil {
log.Ctx(gc.ctx).Warn("IndexCoord remove index on collection fail", zap.Int64("collID", index.CollectionID),
zap.Int64("indexID", index.IndexID), zap.Error(err))
continue
}
} else {
for _, buildID := range buildIDs {
segIdx, ok := gc.metaTable.GetMeta(buildID)
if !ok {
log.Ctx(gc.ctx).Debug("IndexCoord get segment index is not exist", zap.Int64("buildID", buildID))
continue
}
if segIdx.NodeID != 0 {
// wait for releasing reference lock
continue
}
if err := gc.metaTable.RemoveSegmentIndex(segIdx.CollectionID, segIdx.PartitionID, segIdx.SegmentID, segIdx.BuildID); err != nil {
log.Ctx(gc.ctx).Warn("delete index meta from etcd failed, wait to retry", zap.Int64("buildID", segIdx.BuildID),
zap.Int64("nodeID", segIdx.NodeID), zap.Error(err))
continue
}
log.Ctx(gc.ctx).Info("IndexCoord remove segment index meta success", zap.Int64("buildID", segIdx.BuildID),
zap.Int64("nodeID", segIdx.NodeID))
}
log.Ctx(gc.ctx).Info("garbageCollector remove index success", zap.Int64("collID", index.CollectionID),
zap.Int64("indexID", index.IndexID))
}
}
}
}
}
func (gc *garbageCollector) recycleSegIndexesMeta() {
segIndexes := gc.metaTable.GetAllSegIndexes()
collID2segID := make(map[int64]map[int64]struct{})
for segID, segIdx := range segIndexes {
if _, ok := collID2segID[segIdx.CollectionID]; !ok {
collID2segID[segIdx.CollectionID] = make(map[int64]struct{})
}
collID2segID[segIdx.CollectionID][segID] = struct{}{}
}
for collID, segIDs := range collID2segID {
resp, err := gc.indexCoordClient.dataCoordClient.GetFlushedSegments(gc.ctx, &datapb.GetFlushedSegmentsRequest{
CollectionID: collID,
PartitionID: -1,
})
if err != nil {
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector get flushed segments from DataCoord fail",
zap.Int64("collID", collID), zap.Error(err))
return
}
if resp.Status.ErrorCode != commonpb.ErrorCode_Success {
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector get flushed segments from DataCoord fail", zap.Int64("collID", collID),
zap.String("fail reason", resp.Status.Reason))
return
}
flushedSegments := make(map[int64]struct{})
for _, segID := range resp.Segments {
flushedSegments[segID] = struct{}{}
}
for segID := range segIDs {
if segIndexes[segID].IsDeleted {
continue
}
if _, ok := flushedSegments[segID]; !ok {
log.Ctx(gc.ctx).Debug("segment is already not exist, mark it deleted", zap.Int64("collID", collID),
zap.Int64("segID", segID))
if err := gc.metaTable.MarkSegmentsIndexAsDeleted(func(segIndex *model.SegmentIndex) bool {
return segIndex.SegmentID == segID
}); err != nil {
continue
}
}
}
}
//segIndexes := gc.metaTable.GetDeletedSegmentIndexes()
for _, meta := range segIndexes {
if meta.IsDeleted || gc.metaTable.IsIndexDeleted(meta.CollectionID, meta.IndexID) {
if meta.NodeID != 0 {
// wait for releasing reference lock
continue
}
if err := gc.metaTable.RemoveSegmentIndex(meta.CollectionID, meta.PartitionID, meta.SegmentID, meta.BuildID); err != nil {
log.Ctx(gc.ctx).Warn("delete index meta from etcd failed, wait to retry", zap.Int64("buildID", meta.BuildID),
zap.Int64("nodeID", meta.NodeID), zap.Error(err))
continue
}
log.Ctx(gc.ctx).Debug("index meta recycle success", zap.Int64("buildID", meta.BuildID))
}
}
}
func (gc *garbageCollector) recycleUnusedSegIndexes() {
defer gc.wg.Done()
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector recycleUnusedSegIndexes start")
ticker := time.NewTicker(gc.gcMetaDuration)
defer ticker.Stop()
for {
select {
case <-gc.ctx.Done():
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector recycleUnusedMetaLoop context has done")
return
case <-ticker.C:
gc.recycleSegIndexesMeta()
}
}
}
// recycleUnusedIndexFiles is used to delete those index files that no longer exist in the meta.
func (gc *garbageCollector) recycleUnusedIndexFiles() {
defer gc.wg.Done()
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector start recycleUnusedIndexFiles loop")
ticker := time.NewTicker(gc.gcFileDuration)
defer ticker.Stop()
for {
select {
case <-gc.ctx.Done():
return
case <-ticker.C:
prefix := path.Join(gc.chunkManager.RootPath(), common.SegmentIndexPath) + "/"
// list dir first
keys, _, err := gc.chunkManager.ListWithPrefix(gc.ctx, prefix, false)
if err != nil {
log.Ctx(gc.ctx).Error("IndexCoord garbageCollector recycleUnusedIndexFiles list keys from chunk manager failed", zap.Error(err))
continue
}
for _, key := range keys {
log.Ctx(gc.ctx).Debug("indexFiles keys", zap.String("key", key))
buildID, err := parseBuildIDFromFilePath(key)
if err != nil {
log.Ctx(gc.ctx).Error("IndexCoord garbageCollector recycleUnusedIndexFiles parseIndexFileKey", zap.String("key", key), zap.Error(err))
continue
}
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector will recycle index files", zap.Int64("buildID", buildID))
if !gc.metaTable.HasBuildID(buildID) {
// buildID no longer exists in meta, remove all index files
log.Ctx(gc.ctx).Info("IndexCoord garbageCollector recycleUnusedIndexFiles find meta has not exist, remove index files",
zap.Int64("buildID", buildID))
err = gc.chunkManager.RemoveWithPrefix(gc.ctx, key)
if err != nil {
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector recycleUnusedIndexFiles remove index files failed",
zap.Int64("buildID", buildID), zap.String("prefix", key), zap.Error(err))
continue
}
continue
}
log.Ctx(gc.ctx).Info("index meta can be recycled, recycle index files", zap.Int64("buildID", buildID))
canRecycle, indexFilePaths := gc.metaTable.GetIndexFilePathByBuildID(buildID)
if !canRecycle {
// Even if the index is marked as deleted, the index file will not be recycled, wait for the next gc,
// and delete all index files about the buildID at one time.
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector can not recycle index files", zap.Int64("buildID", buildID))
continue
}
filesMap := make(map[string]bool)
for _, file := range indexFilePaths {
filesMap[file] = true
}
files, _, err := gc.chunkManager.ListWithPrefix(gc.ctx, key, true)
if err != nil {
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector recycleUnusedIndexFiles list files failed",
zap.Int64("buildID", buildID), zap.String("prefix", key), zap.Error(err))
continue
}
log.Ctx(gc.ctx).Info("recycle index files", zap.Int64("buildID", buildID), zap.Int("meta files num", len(filesMap)),
zap.Int("chunkManager files num", len(files)))
deletedFilesNum := 0
for _, file := range files {
if _, ok := filesMap[file]; !ok {
if err = gc.chunkManager.Remove(gc.ctx, file); err != nil {
log.Ctx(gc.ctx).Warn("IndexCoord garbageCollector recycleUnusedIndexFiles remove file failed",
zap.Int64("buildID", buildID), zap.String("file", file), zap.Error(err))
continue
}
deletedFilesNum++
}
}
log.Ctx(gc.ctx).Info("index files recycle success", zap.Int64("buildID", buildID),
zap.Int("delete index files num", deletedFilesNum))
}
}
}
}