Files
wehub-resource-sync 498b235461
Build and test / Build and test AMD64 Ubuntu 22.04 (push) Failing after 0s
Publish Builder / amazonlinux2023 (push) Failing after 1s
Build and test / UT for Go (push) Has been skipped
Publish KRTE Images / KRTE (push) Failing after 1s
Build and test / Integration Test (push) Has been skipped
Build and test / Upload Code Coverage (push) Has been skipped
Publish Builder / rockylinux9 (push) Failing after 1s
Publish Builder / ubuntu22.04 (push) Failing after 0s
Publish Builder / ubuntu24.04 (push) Failing after 0s
Publish Gpu Builder / publish-gpu-builder (push) Failing after 1s
Publish Test Images / PyTest (push) Failing after 0s
Build and test / UT for Cpp (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 12:31:17 +08:00

236 lines
7.0 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"
"io"
"runtime/debug"
"time"
"github.com/cockroachdb/errors"
"github.com/samber/lo"
"github.com/milvus-io/milvus-proto/go-api/v3/commonpb"
"github.com/milvus-io/milvus-proto/go-api/v3/schemapb"
"github.com/milvus-io/milvus/internal/storage"
"github.com/milvus-io/milvus/internal/util/importutilv2"
"github.com/milvus-io/milvus/pkg/v3/mlog"
"github.com/milvus-io/milvus/pkg/v3/proto/datapb"
"github.com/milvus-io/milvus/pkg/v3/proto/internalpb"
"github.com/milvus-io/milvus/pkg/v3/util/conc"
"github.com/milvus-io/milvus/pkg/v3/util/merr"
"github.com/milvus-io/milvus/pkg/v3/util/paramtable"
"github.com/milvus-io/milvus/pkg/v3/util/typeutil"
)
type PreImportTask struct {
*datapb.PreImportTask
ctx context.Context
cancel context.CancelFunc
partitionIDs []int64
vchannels []string
schema *schemapb.CollectionSchema
options []*commonpb.KeyValuePair
req *datapb.PreImportRequest
manager TaskManager
cm storage.ChunkManager
}
func NewPreImportTask(req *datapb.PreImportRequest,
manager TaskManager,
cm storage.ChunkManager,
) Task {
fileStats := lo.Map(req.GetImportFiles(), func(file *internalpb.ImportFile, _ int) *datapb.ImportFileStats {
return &datapb.ImportFileStats{
ImportFile: file,
}
})
ctx, cancel := context.WithCancel(context.Background())
// During binlog import, even if the primary key's autoID is set to true,
// the primary key from the binlog should be used instead of being reassigned.
if importutilv2.IsBackup(req.GetOptions()) {
UnsetAutoID(req.GetSchema())
}
return &PreImportTask{
PreImportTask: &datapb.PreImportTask{
JobID: req.GetJobID(),
TaskID: req.GetTaskID(),
CollectionID: req.GetCollectionID(),
State: datapb.ImportTaskStateV2_Pending,
FileStats: fileStats,
},
ctx: ctx,
cancel: cancel,
partitionIDs: req.GetPartitionIDs(),
vchannels: req.GetVchannels(),
schema: req.GetSchema(),
options: req.GetOptions(),
req: req,
manager: manager,
cm: cm,
}
}
func (t *PreImportTask) GetPartitionIDs() []int64 {
return t.partitionIDs
}
func (t *PreImportTask) GetVchannels() []string {
return t.vchannels
}
func (t *PreImportTask) GetType() TaskType {
return PreImportTaskType
}
func (t *PreImportTask) GetSchema() *schemapb.CollectionSchema {
return t.schema
}
func (t *PreImportTask) GetSlots() int64 {
return t.req.GetTaskSlot()
}
// PreImportTask buffer size is fixed
func (t *PreImportTask) GetBufferSize() int64 {
return paramtable.Get().DataNodeCfg.ImportBaseBufferSize.GetAsInt64()
}
func (t *PreImportTask) Cancel() {
t.cancel()
}
func (t *PreImportTask) Clone() Task {
ctx, cancel := context.WithCancel(t.ctx)
return &PreImportTask{
PreImportTask: typeutil.Clone(t.PreImportTask),
ctx: ctx,
cancel: cancel,
partitionIDs: t.GetPartitionIDs(),
vchannels: t.GetVchannels(),
schema: t.GetSchema(),
options: t.options,
req: t.req,
manager: t.manager,
cm: t.cm,
}
}
func (t *PreImportTask) Execute() []*conc.Future[any] {
bufferSize := int(t.GetBufferSize())
mlog.Info(t.ctx, "start to preimport", WrapLogFields(t,
mlog.Int("bufferSize", bufferSize),
mlog.Int64("taskSlot", t.GetSlots()),
mlog.Any("files", t.req.GetImportFiles()),
mlog.Any("schema", t.GetSchema()),
)...)
t.manager.Update(t.GetTaskID(), UpdateState(datapb.ImportTaskStateV2_InProgress))
files := lo.Map(t.GetFileStats(),
func(fileStat *datapb.ImportFileStats, _ int) *internalpb.ImportFile {
return fileStat.GetImportFile()
})
fn := func(i int, file *internalpb.ImportFile) error {
reader, err := importutilv2.NewReader(t.ctx, t.cm, t.GetSchema(), file, t.options, bufferSize, t.req.GetStorageConfig())
if err != nil {
mlog.Warn(t.ctx, "new reader failed", WrapLogFields(t, mlog.String("file", file.String()), mlog.Err(err))...)
reason := fmt.Sprintf("error: %v, file: %s", err, file.String())
t.manager.Update(t.GetTaskID(), UpdateState(datapb.ImportTaskStateV2_Failed), UpdateReason(reason))
return err
}
defer reader.Close()
start := time.Now()
err = t.readFileStat(reader, i)
if err != nil {
mlog.Warn(t.ctx, "preimport failed", WrapLogFields(t, mlog.String("file", file.String()), mlog.Err(err))...)
reason := fmt.Sprintf("error: %v, file: %s", err, file.String())
t.manager.Update(t.GetTaskID(), UpdateState(datapb.ImportTaskStateV2_Failed), UpdateReason(reason))
return err
}
mlog.Info(t.ctx, "read file stat done", WrapLogFields(t, mlog.Strings("files", file.GetPaths()),
mlog.Duration("dur", time.Since(start)))...)
return nil
}
futures := make([]*conc.Future[any], 0, len(files))
for i, file := range files {
i := i
file := file
f := GetExecPool().Submit(func() (any, error) {
defer func() {
debug.FreeOSMemory()
}()
err := fn(i, file)
return err, err
})
futures = append(futures, f)
}
return futures
}
func (t *PreImportTask) readFileStat(reader importutilv2.Reader, fileIdx int) error {
fileSize, err := reader.Size()
if err != nil {
return err
}
maxSize := paramtable.Get().DataNodeCfg.MaxImportFileSizeInGB.GetAsFloat() * 1024 * 1024 * 1024
if fileSize > int64(maxSize) {
return merr.WrapErrParameterInvalidMsg(
"The import file size has reached the maximum limit allowed for importing, "+
"fileSize=%d, maxSize=%d", fileSize, int64(maxSize))
}
totalRows := 0
totalSize := 0
hashedStats := make(map[string]*datapb.PartitionImportStats)
for {
data, err := reader.Read()
if err != nil {
if errors.Is(err, io.EOF) {
break
}
return err
}
err = CheckRowsEqual(t.GetSchema(), data)
if err != nil {
return err
}
rowsCount, err := GetRowsStats(t, data)
if err != nil {
return err
}
MergeHashedStats(rowsCount, hashedStats)
rows := data.GetRowNum()
size := data.GetMemorySize()
totalRows += rows
totalSize += size
mlog.Info(t.ctx, "reading file stat...", WrapLogFields(t, mlog.Int("readRows", rows), mlog.Int("readSize", size))...)
}
stat := &datapb.ImportFileStats{
FileSize: fileSize,
TotalRows: int64(totalRows),
TotalMemorySize: int64(totalSize),
HashedStats: hashedStats,
}
t.manager.Update(t.GetTaskID(), UpdateFileStat(fileIdx, stat))
return nil
}