yihao.dai 7ce876a072
fix: Decoupling importing segment from flush process (#30402)
This pr decoups importing segment from flush process by:
1. Exclude the importing segment from the flush policy, this approch
avoids notifying the datanode to flush the importing segment, which may
not exist.
2. When RootCoord call Flush, DataCoord directly set the importing
segment state to `Flushed`.

issue: https://github.com/milvus-io/milvus/issues/30359

---------

Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
2024-02-03 13:01:12 +08:00

214 lines
7.3 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 importv2
import (
"context"
"fmt"
"github.com/samber/lo"
"go.uber.org/zap"
"github.com/milvus-io/milvus-proto/go-api/v2/commonpb"
"github.com/milvus-io/milvus-proto/go-api/v2/schemapb"
"github.com/milvus-io/milvus/internal/datanode/metacache"
"github.com/milvus-io/milvus/internal/datanode/syncmgr"
"github.com/milvus-io/milvus/internal/proto/datapb"
"github.com/milvus-io/milvus/internal/proto/internalpb"
"github.com/milvus-io/milvus/internal/querycoordv2/params"
"github.com/milvus-io/milvus/internal/storage"
"github.com/milvus-io/milvus/pkg/common"
"github.com/milvus-io/milvus/pkg/log"
"github.com/milvus-io/milvus/pkg/util/merr"
"github.com/milvus-io/milvus/pkg/util/typeutil"
)
func WrapNoTaskError(taskID int64, taskType TaskType) error {
return merr.WrapErrImportFailed(fmt.Sprintf("cannot find %s with id %d", taskType.String(), taskID))
}
func NewSyncTask(ctx context.Context, task *ImportTask, segmentID, partitionID int64, vchannel string, insertData *storage.InsertData) (syncmgr.Task, error) {
if params.Params.CommonCfg.EnableStorageV2.GetAsBool() {
return nil, merr.WrapErrImportFailed("storage v2 is not supported") // TODO: dyh, resolve storage v2
}
metaCache := task.metaCaches[vchannel]
if _, ok := metaCache.GetSegmentByID(segmentID); !ok {
metaCache.AddSegment(&datapb.SegmentInfo{
ID: segmentID,
State: commonpb.SegmentState_Importing,
CollectionID: task.GetCollectionID(),
PartitionID: partitionID,
InsertChannel: vchannel,
}, func(info *datapb.SegmentInfo) *metacache.BloomFilterSet {
bfs := metacache.NewBloomFilterSet()
return bfs
})
}
var serializer syncmgr.Serializer
var err error
serializer, err = syncmgr.NewStorageSerializer(
metaCache,
nil,
)
if err != nil {
return nil, err
}
syncPack := &syncmgr.SyncPack{}
syncPack.WithInsertData(insertData).
WithCollectionID(task.GetCollectionID()).
WithPartitionID(partitionID).
WithChannelName(vchannel).
WithSegmentID(segmentID).
WithTimeRange(task.req.GetTs(), task.req.GetTs()).
WithBatchSize(int64(insertData.GetRowNum()))
return serializer.EncodeBuffer(ctx, syncPack)
}
func NewImportSegmentInfo(syncTask syncmgr.Task, task *ImportTask) (*datapb.ImportSegmentInfo, error) {
segmentID := syncTask.SegmentID()
insertBinlogs, statsBinlog, _ := syncTask.(*syncmgr.SyncTask).Binlogs()
metaCache := task.metaCaches[syncTask.ChannelName()]
segment, ok := metaCache.GetSegmentByID(segmentID)
if !ok {
return nil, merr.WrapErrSegmentNotFound(segmentID, "import failed")
}
return &datapb.ImportSegmentInfo{
SegmentID: segmentID,
ImportedRows: segment.FlushedRows(),
Binlogs: lo.Values(insertBinlogs),
Statslogs: lo.Values(statsBinlog),
}, nil
}
func PickSegment(task *ImportTask, vchannel string, partitionID int64, rows int) int64 {
candidates := lo.Filter(task.req.GetRequestSegments(), func(info *datapb.ImportRequestSegment, _ int) bool {
return info.GetVchannel() == vchannel && info.GetPartitionID() == partitionID
})
importedSegments := lo.KeyBy(task.GetSegmentsInfo(), func(segment *datapb.ImportSegmentInfo) int64 {
return segment.GetSegmentID()
})
for _, candidate := range candidates {
var importedRows int64 = 0
if segment, ok := importedSegments[candidate.GetSegmentID()]; ok {
importedRows = segment.GetImportedRows()
}
if importedRows+int64(rows) <= candidate.GetMaxRows() {
return candidate.GetSegmentID()
}
}
segmentID := lo.MinBy(task.GetSegmentsInfo(), func(s1, s2 *datapb.ImportSegmentInfo) bool {
return s1.GetImportedRows() < s2.GetImportedRows()
}).GetSegmentID()
log.Warn("failed to pick an appropriate segment, opt for the smallest one instead",
WrapLogFields(task, zap.Int64("segmentID", segmentID), zap.Int64("maxRows", candidates[0].GetMaxRows()),
zap.Int("rows", rows), zap.Int64("importedRows", importedSegments[segmentID].GetImportedRows()))...)
return segmentID
}
func CheckRowsEqual(schema *schemapb.CollectionSchema, data *storage.InsertData) error {
if len(data.Data) == 0 {
return nil
}
idToField := lo.KeyBy(schema.GetFields(), func(field *schemapb.FieldSchema) int64 {
return field.GetFieldID()
})
var field int64
var rows int
for fieldID, d := range data.Data {
if idToField[fieldID].GetIsPrimaryKey() && idToField[fieldID].GetAutoID() {
continue
}
field, rows = fieldID, d.RowNum()
break
}
for fieldID, d := range data.Data {
if idToField[fieldID].GetIsPrimaryKey() && idToField[fieldID].GetAutoID() {
continue
}
if d.RowNum() != rows {
return merr.WrapErrImportFailed(
fmt.Sprintf("imported rows are not aligned, field '%s' with '%d' rows, field '%s' with '%d' rows",
idToField[field].GetName(), rows, idToField[fieldID].GetName(), d.RowNum()))
}
}
return nil
}
func AppendSystemFieldsData(task *ImportTask, data *storage.InsertData) error {
idRange := task.req.GetAutoIDRange()
pkField, err := typeutil.GetPrimaryFieldSchema(task.GetSchema())
if err != nil {
return err
}
rowNum := GetInsertDataRowCount(data, task.GetSchema())
ids := make([]int64, rowNum)
for i := 0; i < rowNum; i++ {
ids[i] = idRange.GetBegin() + int64(i)
}
idRange.Begin += int64(rowNum)
if pkField.GetAutoID() {
data.Data[pkField.GetFieldID()] = &storage.Int64FieldData{Data: ids}
}
data.Data[common.RowIDField] = &storage.Int64FieldData{Data: ids}
tss := make([]int64, rowNum)
ts := int64(task.req.GetTs())
for i := 0; i < rowNum; i++ {
tss[i] = ts
}
data.Data[common.TimeStampField] = &storage.Int64FieldData{Data: tss}
return nil
}
func GetInsertDataRowCount(data *storage.InsertData, schema *schemapb.CollectionSchema) int {
fields := lo.KeyBy(schema.GetFields(), func(field *schemapb.FieldSchema) int64 {
return field.GetFieldID()
})
for fieldID, fd := range data.Data {
if fields[fieldID].GetIsDynamic() {
continue
}
if fd.RowNum() != 0 {
return fd.RowNum()
}
}
return 0
}
func LogStats(manager TaskManager) {
logFunc := func(tasks []Task, taskType TaskType) {
byState := lo.GroupBy(tasks, func(t Task) internalpb.ImportState {
return t.GetState()
})
log.Info("import task stats", zap.String("type", taskType.String()),
zap.Int("pending", len(byState[internalpb.ImportState_Pending])),
zap.Int("inProgress", len(byState[internalpb.ImportState_InProgress])),
zap.Int("completed", len(byState[internalpb.ImportState_Completed])),
zap.Int("failed", len(byState[internalpb.ImportState_Failed])))
}
tasks := manager.GetBy(WithType(PreImportTaskType))
logFunc(tasks, PreImportTaskType)
tasks = manager.GetBy(WithType(ImportTaskType))
logFunc(tasks, ImportTaskType)
}