mirror of
https://gitee.com/milvus-io/milvus.git
synced 2024-11-29 18:38:44 +08:00
1ed686783f
Related to #35303 Reduce temporary memory usage for PK interface for segment delete. Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
1495 lines
50 KiB
Go
1495 lines
50 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 querynodev2
|
||
|
||
import (
|
||
"context"
|
||
"fmt"
|
||
"strconv"
|
||
"sync"
|
||
"time"
|
||
|
||
"github.com/samber/lo"
|
||
"go.opentelemetry.io/otel/trace"
|
||
"go.uber.org/zap"
|
||
"golang.org/x/sync/errgroup"
|
||
"google.golang.org/protobuf/proto"
|
||
|
||
"github.com/milvus-io/milvus-proto/go-api/v2/commonpb"
|
||
"github.com/milvus-io/milvus-proto/go-api/v2/milvuspb"
|
||
"github.com/milvus-io/milvus-proto/go-api/v2/msgpb"
|
||
"github.com/milvus-io/milvus-proto/go-api/v2/schemapb"
|
||
"github.com/milvus-io/milvus/internal/proto/datapb"
|
||
"github.com/milvus-io/milvus/internal/proto/indexpb"
|
||
"github.com/milvus-io/milvus/internal/proto/internalpb"
|
||
"github.com/milvus-io/milvus/internal/proto/querypb"
|
||
"github.com/milvus-io/milvus/internal/proto/segcorepb"
|
||
"github.com/milvus-io/milvus/internal/querynodev2/delegator"
|
||
"github.com/milvus-io/milvus/internal/querynodev2/segments"
|
||
"github.com/milvus-io/milvus/internal/querynodev2/tasks"
|
||
"github.com/milvus-io/milvus/internal/storage"
|
||
"github.com/milvus-io/milvus/internal/util/searchutil/scheduler"
|
||
"github.com/milvus-io/milvus/internal/util/streamrpc"
|
||
"github.com/milvus-io/milvus/pkg/common"
|
||
"github.com/milvus-io/milvus/pkg/log"
|
||
"github.com/milvus-io/milvus/pkg/metrics"
|
||
"github.com/milvus-io/milvus/pkg/util/commonpbutil"
|
||
"github.com/milvus-io/milvus/pkg/util/conc"
|
||
"github.com/milvus-io/milvus/pkg/util/funcutil"
|
||
"github.com/milvus-io/milvus/pkg/util/hardware"
|
||
"github.com/milvus-io/milvus/pkg/util/merr"
|
||
"github.com/milvus-io/milvus/pkg/util/metricsinfo"
|
||
"github.com/milvus-io/milvus/pkg/util/paramtable"
|
||
"github.com/milvus-io/milvus/pkg/util/timerecord"
|
||
"github.com/milvus-io/milvus/pkg/util/typeutil"
|
||
)
|
||
|
||
// GetComponentStates returns information about whether the node is healthy
|
||
func (node *QueryNode) GetComponentStates(ctx context.Context, req *milvuspb.GetComponentStatesRequest) (*milvuspb.ComponentStates, error) {
|
||
stats := &milvuspb.ComponentStates{
|
||
Status: merr.Success(),
|
||
}
|
||
|
||
code := node.lifetime.GetState()
|
||
nodeID := common.NotRegisteredID
|
||
|
||
if node.session != nil && node.session.Registered() {
|
||
nodeID = node.GetNodeID()
|
||
}
|
||
log.Debug("QueryNode current state", zap.Int64("NodeID", nodeID), zap.String("StateCode", code.String()))
|
||
|
||
info := &milvuspb.ComponentInfo{
|
||
NodeID: nodeID,
|
||
Role: typeutil.QueryNodeRole,
|
||
StateCode: code,
|
||
}
|
||
stats.State = info
|
||
return stats, nil
|
||
}
|
||
|
||
// GetTimeTickChannel returns the time tick channel
|
||
// TimeTickChannel contains many time tick messages, which will be sent by query nodes
|
||
func (node *QueryNode) GetTimeTickChannel(ctx context.Context, req *internalpb.GetTimeTickChannelRequest) (*milvuspb.StringResponse, error) {
|
||
return &milvuspb.StringResponse{
|
||
Status: merr.Success(),
|
||
Value: paramtable.Get().CommonCfg.QueryCoordTimeTick.GetValue(),
|
||
}, nil
|
||
}
|
||
|
||
// GetStatisticsChannel returns the statistics channel
|
||
// Statistics channel contains statistics infos of query nodes, such as segment infos, memory infos
|
||
func (node *QueryNode) GetStatisticsChannel(ctx context.Context, req *internalpb.GetStatisticsChannelRequest) (*milvuspb.StringResponse, error) {
|
||
return &milvuspb.StringResponse{
|
||
Status: merr.Success(),
|
||
}, nil
|
||
}
|
||
|
||
// GetStatistics returns loaded statistics of collection.
|
||
func (node *QueryNode) GetStatistics(ctx context.Context, req *querypb.GetStatisticsRequest) (*internalpb.GetStatisticsResponse, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("msgID", req.GetReq().GetBase().GetMsgID()),
|
||
zap.Strings("vChannels", req.GetDmlChannels()),
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
zap.Uint64("guaranteeTimestamp", req.GetReq().GetGuaranteeTimestamp()),
|
||
zap.Uint64("timeTravel", req.GetReq().GetTravelTimestamp()),
|
||
)
|
||
log.Debug("received GetStatisticsRequest")
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return &internalpb.GetStatisticsResponse{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
failRet := &internalpb.GetStatisticsResponse{
|
||
Status: merr.Success(),
|
||
}
|
||
|
||
var toReduceResults []*internalpb.GetStatisticsResponse
|
||
var mu sync.Mutex
|
||
runningGp, runningCtx := errgroup.WithContext(ctx)
|
||
for _, ch := range req.GetDmlChannels() {
|
||
ch := ch
|
||
req := &querypb.GetStatisticsRequest{
|
||
Req: req.Req,
|
||
DmlChannels: []string{ch},
|
||
SegmentIDs: req.SegmentIDs,
|
||
FromShardLeader: req.FromShardLeader,
|
||
Scope: req.Scope,
|
||
}
|
||
runningGp.Go(func() error {
|
||
ret, err := node.getChannelStatistics(runningCtx, req, ch)
|
||
if err == nil {
|
||
err = merr.Error(ret.GetStatus())
|
||
}
|
||
|
||
mu.Lock()
|
||
defer mu.Unlock()
|
||
if err != nil {
|
||
failRet.Status = merr.Status(err)
|
||
return err
|
||
}
|
||
toReduceResults = append(toReduceResults, ret)
|
||
return nil
|
||
})
|
||
}
|
||
if err := runningGp.Wait(); err != nil {
|
||
return failRet, nil
|
||
}
|
||
|
||
ret, err := reduceStatisticResponse(toReduceResults)
|
||
if err != nil {
|
||
failRet.Status = merr.Status(err)
|
||
return failRet, nil
|
||
}
|
||
log.Debug("reduce statistic result done")
|
||
|
||
return ret, nil
|
||
}
|
||
|
||
func (node *QueryNode) composeIndexMeta(indexInfos []*indexpb.IndexInfo, schema *schemapb.CollectionSchema) *segcorepb.CollectionIndexMeta {
|
||
fieldIndexMetas := make([]*segcorepb.FieldIndexMeta, 0)
|
||
for _, info := range indexInfos {
|
||
fieldIndexMetas = append(fieldIndexMetas, &segcorepb.FieldIndexMeta{
|
||
CollectionID: info.GetCollectionID(),
|
||
FieldID: info.GetFieldID(),
|
||
IndexName: info.GetIndexName(),
|
||
TypeParams: info.GetTypeParams(),
|
||
IndexParams: info.GetIndexParams(),
|
||
IsAutoIndex: info.GetIsAutoIndex(),
|
||
UserIndexParams: info.GetUserIndexParams(),
|
||
})
|
||
}
|
||
sizePerRecord, err := typeutil.EstimateSizePerRecord(schema)
|
||
maxIndexRecordPerSegment := int64(0)
|
||
if err != nil || sizePerRecord == 0 {
|
||
log.Warn("failed to transfer segment size to collection, because failed to estimate size per record", zap.Error(err))
|
||
} else {
|
||
threshold := paramtable.Get().DataCoordCfg.SegmentMaxSize.GetAsFloat() * 1024 * 1024
|
||
proportion := paramtable.Get().DataCoordCfg.SegmentSealProportion.GetAsFloat()
|
||
maxIndexRecordPerSegment = int64(threshold * proportion / float64(sizePerRecord))
|
||
}
|
||
|
||
return &segcorepb.CollectionIndexMeta{
|
||
IndexMetas: fieldIndexMetas,
|
||
MaxIndexRowCount: maxIndexRecordPerSegment,
|
||
}
|
||
}
|
||
|
||
// WatchDmChannels create consumers on dmChannels to receive Incremental data，which is the important part of real-time query
|
||
func (node *QueryNode) WatchDmChannels(ctx context.Context, req *querypb.WatchDmChannelsRequest) (status *commonpb.Status, e error) {
|
||
defer node.updateDistributionModifyTS()
|
||
|
||
channel := req.GetInfos()[0]
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionID()),
|
||
zap.String("channel", channel.GetChannelName()),
|
||
zap.Int64("currentNodeID", node.GetNodeID()),
|
||
)
|
||
|
||
log.Info("received watch channel request",
|
||
zap.Int64("version", req.GetVersion()),
|
||
)
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
// check index
|
||
if len(req.GetIndexInfoList()) == 0 {
|
||
err := merr.WrapErrIndexNotFoundForCollection(req.GetSchema().GetName())
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
if !node.subscribingChannels.Insert(channel.GetChannelName()) {
|
||
msg := "channel subscribing..."
|
||
log.Warn(msg)
|
||
return merr.Success(), nil
|
||
}
|
||
defer node.subscribingChannels.Remove(channel.GetChannelName())
|
||
|
||
// to avoid concurrent watch/unwatch
|
||
if node.unsubscribingChannels.Contain(channel.GetChannelName()) {
|
||
err := merr.WrapErrChannelReduplicate(channel.GetChannelName(), "the other same channel is unsubscribing")
|
||
log.Warn("failed to unsubscribe channel", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
_, exist := node.delegators.Get(channel.GetChannelName())
|
||
if exist {
|
||
log.Info("channel already subscribed")
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
node.manager.Collection.PutOrRef(req.GetCollectionID(), req.GetSchema(),
|
||
node.composeIndexMeta(req.GetIndexInfoList(), req.Schema), req.GetLoadMeta())
|
||
defer func() {
|
||
if !merr.Ok(status) {
|
||
node.manager.Collection.Unref(req.GetCollectionID(), 1)
|
||
}
|
||
}()
|
||
|
||
delegator, err := delegator.NewShardDelegator(
|
||
ctx,
|
||
req.GetCollectionID(),
|
||
req.GetReplicaID(),
|
||
channel.GetChannelName(),
|
||
req.GetVersion(),
|
||
node.clusterManager,
|
||
node.manager,
|
||
node.tSafeManager,
|
||
node.loader,
|
||
node.factory,
|
||
channel.GetSeekPosition().GetTimestamp(),
|
||
node.queryHook,
|
||
node.chunkManager,
|
||
)
|
||
if err != nil {
|
||
log.Warn("failed to create shard delegator", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
node.delegators.Insert(channel.GetChannelName(), delegator)
|
||
defer func() {
|
||
if err != nil {
|
||
node.delegators.GetAndRemove(channel.GetChannelName())
|
||
}
|
||
}()
|
||
|
||
// create tSafe
|
||
node.tSafeManager.Add(ctx, channel.ChannelName, channel.GetSeekPosition().GetTimestamp())
|
||
defer func() {
|
||
if err != nil {
|
||
node.tSafeManager.Remove(ctx, channel.ChannelName)
|
||
}
|
||
}()
|
||
|
||
pipeline, err := node.pipelineManager.Add(req.GetCollectionID(), channel.GetChannelName())
|
||
if err != nil {
|
||
msg := "failed to create pipeline"
|
||
log.Warn(msg, zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
defer func() {
|
||
if err != nil {
|
||
node.pipelineManager.Remove(channel.GetChannelName())
|
||
}
|
||
}()
|
||
|
||
growingInfo := lo.SliceToMap(channel.GetUnflushedSegmentIds(), func(id int64) (int64, uint64) {
|
||
info := req.GetSegmentInfos()[id]
|
||
return id, info.GetDmlPosition().GetTimestamp()
|
||
})
|
||
delegator.AddExcludedSegments(growingInfo)
|
||
|
||
defer func() {
|
||
if err != nil {
|
||
// remove legacy growing
|
||
node.manager.Segment.RemoveBy(ctx, segments.WithChannel(channel.GetChannelName()),
|
||
segments.WithType(segments.SegmentTypeGrowing))
|
||
// remove legacy l0 segments
|
||
node.manager.Segment.RemoveBy(ctx, segments.WithChannel(channel.GetChannelName()),
|
||
segments.WithLevel(datapb.SegmentLevel_L0))
|
||
}
|
||
}()
|
||
|
||
err = loadL0Segments(ctx, delegator, req)
|
||
if err != nil {
|
||
log.Warn("failed to load l0 segments", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
err = loadGrowingSegments(ctx, delegator, req)
|
||
if err != nil {
|
||
msg := "failed to load growing segments"
|
||
log.Warn(msg, zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
position := &msgpb.MsgPosition{
|
||
ChannelName: channel.SeekPosition.ChannelName,
|
||
MsgID: channel.SeekPosition.MsgID,
|
||
Timestamp: channel.SeekPosition.Timestamp,
|
||
}
|
||
err = pipeline.ConsumeMsgStream(ctx, position)
|
||
if err != nil {
|
||
err = merr.WrapErrServiceUnavailable(err.Error(), "InitPipelineFailed")
|
||
log.Warn(err.Error(),
|
||
zap.Int64("collectionID", channel.CollectionID),
|
||
zap.String("channel", channel.ChannelName),
|
||
)
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
// start pipeline
|
||
pipeline.Start()
|
||
// delegator after all steps done
|
||
delegator.Start()
|
||
log.Info("watch dml channel success")
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
func (node *QueryNode) UnsubDmChannel(ctx context.Context, req *querypb.UnsubDmChannelRequest) (*commonpb.Status, error) {
|
||
defer node.updateDistributionModifyTS()
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionID()),
|
||
zap.String("channel", req.GetChannelName()),
|
||
zap.Int64("currentNodeID", node.GetNodeID()),
|
||
)
|
||
|
||
log.Info("received unsubscribe channel request")
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
node.unsubscribingChannels.Insert(req.GetChannelName())
|
||
defer node.unsubscribingChannels.Remove(req.GetChannelName())
|
||
delegator, ok := node.delegators.GetAndRemove(req.GetChannelName())
|
||
if ok {
|
||
// close the delegator first to block all coming query/search requests
|
||
delegator.Close()
|
||
|
||
node.pipelineManager.Remove(req.GetChannelName())
|
||
node.manager.Segment.RemoveBy(ctx, segments.WithChannel(req.GetChannelName()), segments.WithType(segments.SegmentTypeGrowing))
|
||
_, sealed := node.manager.Segment.RemoveBy(ctx, segments.WithChannel(req.GetChannelName()), segments.WithLevel(datapb.SegmentLevel_L0))
|
||
node.tSafeManager.Remove(ctx, req.GetChannelName())
|
||
|
||
node.manager.Collection.Unref(req.GetCollectionID(), uint32(1+sealed))
|
||
}
|
||
log.Info("unsubscribed channel")
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
func (node *QueryNode) LoadPartitions(ctx context.Context, req *querypb.LoadPartitionsRequest) (*commonpb.Status, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionID()),
|
||
zap.Int64s("partitionIDs", req.GetPartitionIDs()),
|
||
)
|
||
|
||
log.Info("received load partitions request")
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthyOrStopping); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
collection := node.manager.Collection.Get(req.GetCollectionID())
|
||
if collection != nil {
|
||
collection.AddPartition(req.GetPartitionIDs()...)
|
||
}
|
||
|
||
log.Info("load partitions done")
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// LoadSegments load historical data into query node, historical data can be vector data or index
|
||
func (node *QueryNode) LoadSegments(ctx context.Context, req *querypb.LoadSegmentsRequest) (*commonpb.Status, error) {
|
||
defer node.updateDistributionModifyTS()
|
||
segment := req.GetInfos()[0]
|
||
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", segment.GetCollectionID()),
|
||
zap.Int64("partitionID", segment.GetPartitionID()),
|
||
zap.String("shard", segment.GetInsertChannel()),
|
||
zap.Int64("segmentID", segment.GetSegmentID()),
|
||
zap.String("level", segment.GetLevel().String()),
|
||
zap.Int64("currentNodeID", node.GetNodeID()),
|
||
zap.Int64("dstNodeID", req.GetDstNodeID()),
|
||
)
|
||
|
||
log.Info("received load segments request",
|
||
zap.Int64("version", req.GetVersion()),
|
||
zap.Bool("needTransfer", req.GetNeedTransfer()),
|
||
zap.String("loadScope", req.GetLoadScope().String()))
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
// check index
|
||
if len(req.GetIndexInfoList()) == 0 {
|
||
err := merr.WrapErrIndexNotFoundForCollection(req.GetSchema().GetName())
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
// fallback binlog memory size to log size when it is zero
|
||
fallbackBinlogMemorySize := func(binlogs []*datapb.FieldBinlog) {
|
||
for _, insertBinlogs := range binlogs {
|
||
for _, b := range insertBinlogs.GetBinlogs() {
|
||
if b.GetMemorySize() == 0 {
|
||
b.MemorySize = b.GetLogSize()
|
||
}
|
||
}
|
||
}
|
||
}
|
||
for _, s := range req.GetInfos() {
|
||
fallbackBinlogMemorySize(s.GetBinlogPaths())
|
||
fallbackBinlogMemorySize(s.GetStatslogs())
|
||
fallbackBinlogMemorySize(s.GetDeltalogs())
|
||
}
|
||
|
||
// Delegates request to workers
|
||
if req.GetNeedTransfer() {
|
||
delegator, ok := node.delegators.Get(segment.GetInsertChannel())
|
||
if !ok {
|
||
msg := "failed to load segments, delegator not found"
|
||
log.Warn(msg)
|
||
err := merr.WrapErrChannelNotFound(segment.GetInsertChannel())
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
if len(req.GetInfos()) > 0 && req.GetInfos()[0].Level == datapb.SegmentLevel_L0 {
|
||
// force l0 segment to load on delegator
|
||
if req.DstNodeID != node.GetNodeID() {
|
||
log.Info("unexpected L0 segment load on non-delegator node, force to load on delegator")
|
||
req.DstNodeID = node.GetNodeID()
|
||
}
|
||
}
|
||
|
||
req.NeedTransfer = false
|
||
err := delegator.LoadSegments(ctx, req)
|
||
if err != nil {
|
||
log.Warn("delegator failed to load segments", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
node.manager.Collection.PutOrRef(req.GetCollectionID(), req.GetSchema(),
|
||
node.composeIndexMeta(req.GetIndexInfoList(), req.GetSchema()), req.GetLoadMeta())
|
||
defer node.manager.Collection.Unref(req.GetCollectionID(), 1)
|
||
|
||
if req.GetLoadScope() == querypb.LoadScope_Delta {
|
||
return node.loadDeltaLogs(ctx, req), nil
|
||
}
|
||
if req.GetLoadScope() == querypb.LoadScope_Index {
|
||
return node.loadIndex(ctx, req), nil
|
||
}
|
||
|
||
// Actual load segment
|
||
log.Info("start to load segments...")
|
||
loaded, err := node.loader.Load(ctx,
|
||
req.GetCollectionID(),
|
||
segments.SegmentTypeSealed,
|
||
req.GetVersion(),
|
||
req.GetInfos()...,
|
||
)
|
||
if err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
node.manager.Collection.Ref(req.GetCollectionID(), uint32(len(loaded)))
|
||
|
||
log.Info("load segments done...",
|
||
zap.Int64s("segments", lo.Map(loaded, func(s segments.Segment, _ int) int64 { return s.ID() })))
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// ReleaseCollection clears all data related to this collection on the querynode
|
||
func (node *QueryNode) ReleaseCollection(ctx context.Context, in *querypb.ReleaseCollectionRequest) (*commonpb.Status, error) {
|
||
if err := node.lifetime.Add(merr.IsHealthyOrStopping); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// ReleasePartitions clears all data related to this partition on the querynode
|
||
func (node *QueryNode) ReleasePartitions(ctx context.Context, req *querypb.ReleasePartitionsRequest) (*commonpb.Status, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collection", req.GetCollectionID()),
|
||
zap.Int64s("partitions", req.GetPartitionIDs()),
|
||
)
|
||
|
||
log.Info("received release partitions request")
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
collection := node.manager.Collection.Get(req.GetCollectionID())
|
||
if collection != nil {
|
||
for _, partition := range req.GetPartitionIDs() {
|
||
collection.RemovePartition(partition)
|
||
}
|
||
}
|
||
|
||
log.Info("release partitions done")
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// ReleaseSegments remove the specified segments from query node according segmentIDs, partitionIDs, and collectionID
|
||
func (node *QueryNode) ReleaseSegments(ctx context.Context, req *querypb.ReleaseSegmentsRequest) (*commonpb.Status, error) {
|
||
defer node.updateDistributionModifyTS()
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionID()),
|
||
zap.String("shard", req.GetShard()),
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
zap.Int64("currentNodeID", node.GetNodeID()),
|
||
)
|
||
|
||
log.Info("received release segment request",
|
||
zap.String("scope", req.GetScope().String()),
|
||
zap.Bool("needTransfer", req.GetNeedTransfer()),
|
||
)
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthyOrStopping); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
if req.GetNeedTransfer() {
|
||
delegator, ok := node.delegators.Get(req.GetShard())
|
||
if !ok {
|
||
msg := "failed to release segment, delegator not found"
|
||
log.Warn(msg)
|
||
err := merr.WrapErrChannelNotFound(req.GetShard())
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
req.NeedTransfer = false
|
||
err := delegator.ReleaseSegments(ctx, req, false)
|
||
if err != nil {
|
||
log.Warn("delegator failed to release segment", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
log.Info("start to release segments")
|
||
sealedCount := 0
|
||
for _, id := range req.GetSegmentIDs() {
|
||
_, count := node.manager.Segment.Remove(ctx, id, req.GetScope())
|
||
sealedCount += count
|
||
}
|
||
node.manager.Collection.Unref(req.GetCollectionID(), uint32(sealedCount))
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// GetSegmentInfo returns segment information of the collection on the queryNode, and the information includes memSize, numRow, indexName, indexID ...
|
||
func (node *QueryNode) GetSegmentInfo(ctx context.Context, in *querypb.GetSegmentInfoRequest) (*querypb.GetSegmentInfoResponse, error) {
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return &querypb.GetSegmentInfoResponse{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
var segmentInfos []*querypb.SegmentInfo
|
||
for _, segmentID := range in.GetSegmentIDs() {
|
||
segment := node.manager.Segment.Get(segmentID)
|
||
if segment == nil {
|
||
continue
|
||
}
|
||
collection := node.manager.Collection.Get(segment.Collection())
|
||
if collection == nil {
|
||
continue
|
||
}
|
||
|
||
// TODO(yah01): now Milvus supports only 1 vector field
|
||
vecFields := funcutil.GetVecFieldIDs(collection.Schema())
|
||
var (
|
||
indexName string
|
||
indexID int64
|
||
indexInfos []*querypb.FieldIndexInfo
|
||
)
|
||
for _, field := range vecFields {
|
||
index := segment.GetIndex(field)
|
||
if index != nil {
|
||
indexName = index.IndexInfo.GetIndexName()
|
||
indexID = index.IndexInfo.GetIndexID()
|
||
indexInfos = append(indexInfos, index.IndexInfo)
|
||
}
|
||
}
|
||
|
||
info := &querypb.SegmentInfo{
|
||
SegmentID: segment.ID(),
|
||
SegmentState: segment.Type(),
|
||
DmChannel: segment.Shard().VirtualName(),
|
||
PartitionID: segment.Partition(),
|
||
CollectionID: segment.Collection(),
|
||
NodeID: node.GetNodeID(),
|
||
NodeIds: []int64{node.GetNodeID()},
|
||
MemSize: segment.MemSize(),
|
||
NumRows: segment.InsertCount(),
|
||
IndexName: indexName,
|
||
IndexID: indexID,
|
||
IndexInfos: indexInfos,
|
||
}
|
||
segmentInfos = append(segmentInfos, info)
|
||
}
|
||
|
||
return &querypb.GetSegmentInfoResponse{
|
||
Status: merr.Success(),
|
||
Infos: segmentInfos,
|
||
}, nil
|
||
}
|
||
|
||
// only used for shard delegator search segments from worker
|
||
func (node *QueryNode) SearchSegments(ctx context.Context, req *querypb.SearchRequest) (*internalpb.SearchResults, error) {
|
||
channel := req.GetDmlChannels()[0]
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("msgID", req.GetReq().GetBase().GetMsgID()),
|
||
zap.Int64("collectionID", req.Req.GetCollectionID()),
|
||
zap.String("channel", channel),
|
||
zap.String("scope", req.GetScope().String()),
|
||
)
|
||
channelsMvcc := make(map[string]uint64)
|
||
for _, ch := range req.GetDmlChannels() {
|
||
channelsMvcc[ch] = req.GetReq().GetMvccTimestamp()
|
||
}
|
||
resp := &internalpb.SearchResults{
|
||
ChannelsMvcc: channelsMvcc,
|
||
}
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.SearchLabel, metrics.TotalLabel, metrics.FromLeader).Inc()
|
||
defer func() {
|
||
if !merr.Ok(resp.GetStatus()) {
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.SearchLabel, metrics.FailLabel, metrics.FromLeader).Inc()
|
||
}
|
||
}()
|
||
|
||
log.Debug("start to search segments on worker",
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
)
|
||
searchCtx, cancel := context.WithCancel(ctx)
|
||
defer cancel()
|
||
|
||
tr := timerecord.NewTimeRecorder("searchSegments")
|
||
log.Debug("search segments...")
|
||
|
||
collection := node.manager.Collection.Get(req.Req.GetCollectionID())
|
||
if collection == nil {
|
||
err := merr.WrapErrCollectionNotLoaded(req.GetReq().GetCollectionID())
|
||
log.Warn("failed to search segments", zap.Error(err))
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
var task scheduler.Task
|
||
if paramtable.Get().QueryNodeCfg.UseStreamComputing.GetAsBool() {
|
||
task = tasks.NewStreamingSearchTask(searchCtx, collection, node.manager, req, node.serverID)
|
||
} else {
|
||
task = tasks.NewSearchTask(searchCtx, collection, node.manager, req, node.serverID)
|
||
}
|
||
|
||
if err := node.scheduler.Add(task); err != nil {
|
||
log.Warn("failed to search channel", zap.Error(err))
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
err := task.Wait()
|
||
if err != nil {
|
||
log.Warn("failed to search segments", zap.Error(err))
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
tr.CtxElapse(ctx, fmt.Sprintf("search segments done, channel = %s, segmentIDs = %v",
|
||
channel,
|
||
req.GetSegmentIDs(),
|
||
))
|
||
|
||
latency := tr.ElapseSpan()
|
||
metrics.QueryNodeSQReqLatency.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.SearchLabel, metrics.FromLeader).Observe(float64(latency.Milliseconds()))
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.SearchLabel, metrics.SuccessLabel, metrics.FromLeader).Inc()
|
||
|
||
resp = task.SearchResult()
|
||
resp.GetCostAggregation().ResponseTime = tr.ElapseSpan().Milliseconds()
|
||
resp.GetCostAggregation().TotalNQ = node.scheduler.GetWaitingTaskTotalNQ()
|
||
if req.GetReq().GetIsTopkReduce() {
|
||
resp.IsTopkReduce = true
|
||
}
|
||
return resp, nil
|
||
}
|
||
|
||
// Search performs replica search tasks.
|
||
func (node *QueryNode) Search(ctx context.Context, req *querypb.SearchRequest) (*internalpb.SearchResults, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetReq().GetCollectionID()),
|
||
zap.Strings("channels", req.GetDmlChannels()),
|
||
zap.Int64("nq", req.GetReq().GetNq()),
|
||
)
|
||
|
||
log.Debug("Received SearchRequest",
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
zap.Uint64("guaranteeTimestamp", req.GetReq().GetGuaranteeTimestamp()),
|
||
zap.Uint64("mvccTimestamp", req.GetReq().GetMvccTimestamp()))
|
||
|
||
tr := timerecord.NewTimeRecorderWithTrace(ctx, "SearchRequest")
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return &internalpb.SearchResults{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
resp := &internalpb.SearchResults{
|
||
Status: merr.Success(),
|
||
}
|
||
collection := node.manager.Collection.Get(req.GetReq().GetCollectionID())
|
||
if collection == nil {
|
||
resp.Status = merr.Status(merr.WrapErrCollectionNotFound(req.GetReq().GetCollectionID()))
|
||
return resp, nil
|
||
}
|
||
|
||
if len(req.GetDmlChannels()) != 1 {
|
||
err := merr.WrapErrParameterInvalid(1, len(req.GetDmlChannels()), "count of channel to be searched should only be 1, wrong code")
|
||
resp.Status = merr.Status(err)
|
||
log.Warn("got wrong number of channels to be searched", zap.Error(err))
|
||
return resp, nil
|
||
}
|
||
|
||
ch := req.GetDmlChannels()[0]
|
||
channelReq := &querypb.SearchRequest{
|
||
Req: req.Req,
|
||
DmlChannels: []string{ch},
|
||
SegmentIDs: req.SegmentIDs,
|
||
Scope: req.Scope,
|
||
TotalChannelNum: req.TotalChannelNum,
|
||
}
|
||
ret, err := node.searchChannel(ctx, channelReq, ch)
|
||
if err != nil {
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
tr.RecordSpan()
|
||
ret.Status = merr.Success()
|
||
|
||
reduceLatency := tr.RecordSpan()
|
||
metrics.QueryNodeReduceLatency.
|
||
WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.SearchLabel, metrics.ReduceShards, metrics.BatchReduce).
|
||
Observe(float64(reduceLatency.Milliseconds()))
|
||
metrics.QueryNodeExecuteCounter.WithLabelValues(strconv.FormatInt(node.GetNodeID(), 10), metrics.SearchLabel).
|
||
Add(float64(proto.Size(req)))
|
||
|
||
if ret.GetCostAggregation() != nil {
|
||
ret.GetCostAggregation().ResponseTime = tr.ElapseSpan().Milliseconds()
|
||
}
|
||
return ret, nil
|
||
}
|
||
|
||
// only used for delegator query segments from worker
|
||
func (node *QueryNode) QuerySegments(ctx context.Context, req *querypb.QueryRequest) (*internalpb.RetrieveResults, error) {
|
||
resp := &internalpb.RetrieveResults{
|
||
Status: merr.Success(),
|
||
}
|
||
msgID := req.Req.Base.GetMsgID()
|
||
traceID := trace.SpanFromContext(ctx).SpanContext().TraceID()
|
||
channel := req.GetDmlChannels()[0]
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("msgID", msgID),
|
||
zap.Int64("collectionID", req.GetReq().GetCollectionID()),
|
||
zap.String("channel", channel),
|
||
zap.String("scope", req.GetScope().String()),
|
||
)
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.TotalLabel, metrics.FromLeader).Inc()
|
||
defer func() {
|
||
if resp.GetStatus().GetErrorCode() != commonpb.ErrorCode_Success {
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.FailLabel, metrics.FromLeader).Inc()
|
||
}
|
||
}()
|
||
|
||
log.Debug("start do query segments", zap.Int64s("segmentIDs", req.GetSegmentIDs()))
|
||
// add cancel when error occurs
|
||
queryCtx, cancel := context.WithCancel(ctx)
|
||
defer cancel()
|
||
|
||
tr := timerecord.NewTimeRecorder("querySegments")
|
||
collection := node.manager.Collection.Get(req.Req.GetCollectionID())
|
||
if collection == nil {
|
||
resp.Status = merr.Status(merr.WrapErrCollectionNotLoaded(req.Req.GetCollectionID()))
|
||
return resp, nil
|
||
}
|
||
|
||
// Send task to scheduler and wait until it finished.
|
||
task := tasks.NewQueryTask(queryCtx, collection, node.manager, req)
|
||
if err := node.scheduler.Add(task); err != nil {
|
||
log.Warn("failed to add query task into scheduler", zap.Error(err))
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
err := task.Wait()
|
||
if err != nil {
|
||
log.Warn("failed to query channel", zap.Error(err))
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
tr.CtxElapse(ctx, fmt.Sprintf("do query done, traceID = %s, vChannel = %s, segmentIDs = %v",
|
||
traceID,
|
||
channel,
|
||
req.GetSegmentIDs(),
|
||
))
|
||
|
||
// TODO QueryNodeSQLatencyInQueue QueryNodeReduceLatency
|
||
latency := tr.ElapseSpan()
|
||
metrics.QueryNodeSQReqLatency.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.FromLeader).Observe(float64(latency.Milliseconds()))
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.SuccessLabel, metrics.FromLeader).Inc()
|
||
result := task.Result()
|
||
result.GetCostAggregation().ResponseTime = latency.Milliseconds()
|
||
result.GetCostAggregation().TotalNQ = node.scheduler.GetWaitingTaskTotalNQ()
|
||
return result, nil
|
||
}
|
||
|
||
// Query performs replica query tasks.
|
||
func (node *QueryNode) Query(ctx context.Context, req *querypb.QueryRequest) (*internalpb.RetrieveResults, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetReq().GetCollectionID()),
|
||
zap.Strings("shards", req.GetDmlChannels()),
|
||
)
|
||
|
||
log.Debug("received query request",
|
||
zap.Int64s("outputFields", req.GetReq().GetOutputFieldsId()),
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
zap.Uint64("guaranteeTimestamp", req.GetReq().GetGuaranteeTimestamp()),
|
||
zap.Uint64("mvccTimestamp", req.GetReq().GetMvccTimestamp()),
|
||
zap.Bool("isCount", req.GetReq().GetIsCount()),
|
||
)
|
||
tr := timerecord.NewTimeRecorderWithTrace(ctx, "QueryRequest")
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return &internalpb.RetrieveResults{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
toMergeResults := make([]*internalpb.RetrieveResults, len(req.GetDmlChannels()))
|
||
runningGp, runningCtx := errgroup.WithContext(ctx)
|
||
|
||
for i, ch := range req.GetDmlChannels() {
|
||
ch := ch
|
||
req := &querypb.QueryRequest{
|
||
Req: req.Req,
|
||
DmlChannels: []string{ch},
|
||
SegmentIDs: req.SegmentIDs,
|
||
Scope: req.Scope,
|
||
}
|
||
|
||
idx := i
|
||
runningGp.Go(func() error {
|
||
ret, err := node.queryChannel(runningCtx, req, ch)
|
||
if err == nil {
|
||
err = merr.Error(ret.GetStatus())
|
||
}
|
||
if err != nil {
|
||
return err
|
||
}
|
||
toMergeResults[idx] = ret
|
||
return nil
|
||
})
|
||
}
|
||
if err := runningGp.Wait(); err != nil {
|
||
return &internalpb.RetrieveResults{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
|
||
tr.RecordSpan()
|
||
reducer := segments.CreateInternalReducer(req, node.manager.Collection.Get(req.GetReq().GetCollectionID()).Schema())
|
||
ret, err := reducer.Reduce(ctx, toMergeResults)
|
||
if err != nil {
|
||
return &internalpb.RetrieveResults{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
reduceLatency := tr.RecordSpan()
|
||
metrics.QueryNodeReduceLatency.WithLabelValues(fmt.Sprint(node.GetNodeID()),
|
||
metrics.QueryLabel, metrics.ReduceShards, metrics.BatchReduce).
|
||
Observe(float64(reduceLatency.Milliseconds()))
|
||
|
||
metrics.QueryNodeExecuteCounter.WithLabelValues(strconv.FormatInt(node.GetNodeID(), 10), metrics.QueryLabel).Add(float64(proto.Size(req)))
|
||
relatedDataSize := lo.Reduce(toMergeResults, func(acc int64, result *internalpb.RetrieveResults, _ int) int64 {
|
||
return acc + result.GetCostAggregation().GetTotalRelatedDataSize()
|
||
}, 0)
|
||
|
||
if ret.CostAggregation == nil {
|
||
ret.CostAggregation = &internalpb.CostAggregation{}
|
||
}
|
||
ret.CostAggregation.ResponseTime = tr.ElapseSpan().Milliseconds()
|
||
ret.CostAggregation.TotalRelatedDataSize = relatedDataSize
|
||
return ret, nil
|
||
}
|
||
|
||
func (node *QueryNode) QueryStream(req *querypb.QueryRequest, srv querypb.QueryNode_QueryStreamServer) error {
|
||
ctx := srv.Context()
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetReq().GetCollectionID()),
|
||
zap.Strings("shards", req.GetDmlChannels()),
|
||
)
|
||
concurrentSrv := streamrpc.NewConcurrentQueryStreamServer(srv)
|
||
|
||
log.Debug("received query stream request",
|
||
zap.Int64s("outputFields", req.GetReq().GetOutputFieldsId()),
|
||
zap.Int64s("segmentIDs", req.GetSegmentIDs()),
|
||
zap.Uint64("guaranteeTimestamp", req.GetReq().GetGuaranteeTimestamp()),
|
||
zap.Uint64("mvccTimestamp", req.GetReq().GetMvccTimestamp()),
|
||
zap.Bool("isCount", req.GetReq().GetIsCount()),
|
||
)
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
concurrentSrv.Send(&internalpb.RetrieveResults{Status: merr.Status(err)})
|
||
return nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
runningGp, runningCtx := errgroup.WithContext(ctx)
|
||
|
||
for _, ch := range req.GetDmlChannels() {
|
||
ch := ch
|
||
req := &querypb.QueryRequest{
|
||
Req: req.Req,
|
||
DmlChannels: []string{ch},
|
||
SegmentIDs: req.SegmentIDs,
|
||
Scope: req.Scope,
|
||
}
|
||
|
||
runningGp.Go(func() error {
|
||
err := node.queryChannelStream(runningCtx, req, ch, concurrentSrv)
|
||
if err != nil {
|
||
return err
|
||
}
|
||
return nil
|
||
})
|
||
}
|
||
|
||
if err := runningGp.Wait(); err != nil {
|
||
concurrentSrv.Send(&internalpb.RetrieveResults{
|
||
Status: merr.Status(err),
|
||
})
|
||
return nil
|
||
}
|
||
|
||
metrics.QueryNodeExecuteCounter.WithLabelValues(strconv.FormatInt(node.GetNodeID(), 10), metrics.QueryLabel).Add(float64(proto.Size(req)))
|
||
return nil
|
||
}
|
||
|
||
func (node *QueryNode) QueryStreamSegments(req *querypb.QueryRequest, srv querypb.QueryNode_QueryStreamSegmentsServer) error {
|
||
ctx := srv.Context()
|
||
msgID := req.Req.Base.GetMsgID()
|
||
traceID := trace.SpanFromContext(ctx).SpanContext().TraceID()
|
||
channel := req.GetDmlChannels()[0]
|
||
concurrentSrv := streamrpc.NewConcurrentQueryStreamServer(srv)
|
||
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("msgID", msgID),
|
||
zap.Int64("collectionID", req.GetReq().GetCollectionID()),
|
||
zap.String("channel", channel),
|
||
zap.String("scope", req.GetScope().String()),
|
||
)
|
||
|
||
resp := &internalpb.RetrieveResults{}
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.TotalLabel, metrics.FromLeader).Inc()
|
||
defer func() {
|
||
if resp.GetStatus().GetErrorCode() != commonpb.ErrorCode_Success {
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.FailLabel, metrics.FromLeader).Inc()
|
||
}
|
||
}()
|
||
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
resp.Status = merr.Status(err)
|
||
concurrentSrv.Send(resp)
|
||
return nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
log.Debug("start do query with channel", zap.Int64s("segmentIDs", req.GetSegmentIDs()))
|
||
|
||
tr := timerecord.NewTimeRecorder("queryChannel")
|
||
|
||
err := node.queryStreamSegments(ctx, req, concurrentSrv)
|
||
if err != nil {
|
||
resp.Status = merr.Status(err)
|
||
concurrentSrv.Send(resp)
|
||
return nil
|
||
}
|
||
|
||
tr.CtxElapse(ctx, fmt.Sprintf("do query done, traceID = %s, vChannel = %s, segmentIDs = %v",
|
||
traceID,
|
||
channel,
|
||
req.GetSegmentIDs(),
|
||
))
|
||
|
||
// TODO QueryNodeSQLatencyInQueue QueryNodeReduceLatency
|
||
latency := tr.ElapseSpan()
|
||
metrics.QueryNodeSQReqLatency.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.FromLeader).Observe(float64(latency.Milliseconds()))
|
||
metrics.QueryNodeSQCount.WithLabelValues(fmt.Sprint(node.GetNodeID()), metrics.QueryLabel, metrics.SuccessLabel, metrics.FromLeader).Inc()
|
||
return nil
|
||
}
|
||
|
||
// SyncReplicaSegments syncs replica node & segments states
|
||
func (node *QueryNode) SyncReplicaSegments(ctx context.Context, req *querypb.SyncReplicaSegmentsRequest) (*commonpb.Status, error) {
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// ShowConfigurations returns the configurations of queryNode matching req.Pattern
|
||
func (node *QueryNode) ShowConfigurations(ctx context.Context, req *internalpb.ShowConfigurationsRequest) (*internalpb.ShowConfigurationsResponse, error) {
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
log.Warn("QueryNode.ShowConfigurations failed",
|
||
zap.Int64("nodeId", node.GetNodeID()),
|
||
zap.String("req", req.Pattern),
|
||
zap.Error(err))
|
||
|
||
return &internalpb.ShowConfigurationsResponse{
|
||
Status: merr.Status(err),
|
||
Configuations: nil,
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
configList := make([]*commonpb.KeyValuePair, 0)
|
||
for key, value := range paramtable.Get().GetComponentConfigurations("querynode", req.Pattern) {
|
||
configList = append(configList,
|
||
&commonpb.KeyValuePair{
|
||
Key: key,
|
||
Value: value,
|
||
})
|
||
}
|
||
|
||
return &internalpb.ShowConfigurationsResponse{
|
||
Status: merr.Success(),
|
||
Configuations: configList,
|
||
}, nil
|
||
}
|
||
|
||
// GetMetrics return system infos of the query node, such as total memory, memory usage, cpu usage ...
|
||
func (node *QueryNode) GetMetrics(ctx context.Context, req *milvuspb.GetMetricsRequest) (*milvuspb.GetMetricsResponse, error) {
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
log.Warn("QueryNode.GetMetrics failed",
|
||
zap.Int64("nodeId", node.GetNodeID()),
|
||
zap.String("req", req.Request),
|
||
zap.Error(err))
|
||
|
||
return &milvuspb.GetMetricsResponse{
|
||
Status: merr.Status(err),
|
||
Response: "",
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
resp := &milvuspb.GetMetricsResponse{
|
||
Status: merr.Success(),
|
||
ComponentName: metricsinfo.ConstructComponentName(typeutil.QueryNodeRole, paramtable.GetNodeID()),
|
||
}
|
||
|
||
ret, err := node.metricsRequest.ExecuteMetricsRequest(ctx, req)
|
||
if err != nil {
|
||
resp.Status = merr.Status(err)
|
||
return resp, nil
|
||
}
|
||
|
||
resp.Response = ret
|
||
return resp, nil
|
||
}
|
||
|
||
func (node *QueryNode) GetDataDistribution(ctx context.Context, req *querypb.GetDataDistributionRequest) (*querypb.GetDataDistributionResponse, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("msgID", req.GetBase().GetMsgID()),
|
||
zap.Int64("nodeID", node.GetNodeID()),
|
||
)
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
log.Warn("QueryNode.GetDataDistribution failed",
|
||
zap.Error(err))
|
||
|
||
return &querypb.GetDataDistributionResponse{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
lastModifyTs := node.getDistributionModifyTS()
|
||
distributionChange := func() bool {
|
||
if req.GetLastUpdateTs() == 0 {
|
||
return true
|
||
}
|
||
|
||
return req.GetLastUpdateTs() < lastModifyTs
|
||
}
|
||
|
||
if !distributionChange() {
|
||
return &querypb.GetDataDistributionResponse{
|
||
Status: merr.Success(),
|
||
NodeID: node.GetNodeID(),
|
||
LastModifyTs: lastModifyTs,
|
||
}, nil
|
||
}
|
||
|
||
sealedSegments := node.manager.Segment.GetBy(segments.WithType(commonpb.SegmentState_Sealed))
|
||
segmentVersionInfos := make([]*querypb.SegmentVersionInfo, 0, len(sealedSegments))
|
||
for _, s := range sealedSegments {
|
||
segmentVersionInfos = append(segmentVersionInfos, &querypb.SegmentVersionInfo{
|
||
ID: s.ID(),
|
||
Collection: s.Collection(),
|
||
Partition: s.Partition(),
|
||
Channel: s.Shard().VirtualName(),
|
||
Version: s.Version(),
|
||
Level: s.Level(),
|
||
IsSorted: s.IsSorted(),
|
||
LastDeltaTimestamp: s.LastDeltaTimestamp(),
|
||
IndexInfo: lo.SliceToMap(s.Indexes(), func(info *segments.IndexedFieldInfo) (int64, *querypb.FieldIndexInfo) {
|
||
return info.IndexInfo.FieldID, info.IndexInfo
|
||
}),
|
||
})
|
||
}
|
||
|
||
channelVersionInfos := make([]*querypb.ChannelVersionInfo, 0)
|
||
leaderViews := make([]*querypb.LeaderView, 0)
|
||
|
||
node.delegators.Range(func(key string, delegator delegator.ShardDelegator) bool {
|
||
if !delegator.Serviceable() {
|
||
return true
|
||
}
|
||
channelVersionInfos = append(channelVersionInfos, &querypb.ChannelVersionInfo{
|
||
Channel: key,
|
||
Collection: delegator.Collection(),
|
||
Version: delegator.Version(),
|
||
})
|
||
|
||
sealed, growing := delegator.GetSegmentInfo(false)
|
||
sealedSegments := make(map[int64]*querypb.SegmentDist)
|
||
for _, item := range sealed {
|
||
for _, segment := range item.Segments {
|
||
sealedSegments[segment.SegmentID] = &querypb.SegmentDist{
|
||
NodeID: item.NodeID,
|
||
Version: segment.Version,
|
||
}
|
||
}
|
||
}
|
||
|
||
numOfGrowingRows := int64(0)
|
||
growingSegments := make(map[int64]*msgpb.MsgPosition)
|
||
for _, entry := range growing {
|
||
segment := node.manager.Segment.GetWithType(entry.SegmentID, segments.SegmentTypeGrowing)
|
||
if segment == nil {
|
||
log.Warn("leader view growing not found", zap.String("channel", key), zap.Int64("segmentID", entry.SegmentID))
|
||
growingSegments[entry.SegmentID] = &msgpb.MsgPosition{}
|
||
continue
|
||
}
|
||
growingSegments[entry.SegmentID] = segment.StartPosition()
|
||
numOfGrowingRows += segment.InsertCount()
|
||
}
|
||
|
||
leaderViews = append(leaderViews, &querypb.LeaderView{
|
||
Collection: delegator.Collection(),
|
||
Channel: key,
|
||
SegmentDist: sealedSegments,
|
||
GrowingSegments: growingSegments,
|
||
TargetVersion: delegator.GetTargetVersion(),
|
||
NumOfGrowingRows: numOfGrowingRows,
|
||
PartitionStatsVersions: delegator.GetPartitionStatsVersions(ctx),
|
||
})
|
||
return true
|
||
})
|
||
|
||
return &querypb.GetDataDistributionResponse{
|
||
Status: merr.Success(),
|
||
NodeID: node.GetNodeID(),
|
||
Segments: segmentVersionInfos,
|
||
Channels: channelVersionInfos,
|
||
LeaderViews: leaderViews,
|
||
LastModifyTs: lastModifyTs,
|
||
MemCapacityInMB: float64(hardware.GetMemoryCount() / 1024 / 1024),
|
||
}, nil
|
||
}
|
||
|
||
func (node *QueryNode) SyncDistribution(ctx context.Context, req *querypb.SyncDistributionRequest) (*commonpb.Status, error) {
|
||
defer node.updateDistributionModifyTS()
|
||
|
||
log := log.Ctx(ctx).With(zap.Int64("collectionID", req.GetCollectionID()),
|
||
zap.String("channel", req.GetChannel()), zap.Int64("currentNodeID", node.GetNodeID()))
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
// get shard delegator
|
||
shardDelegator, ok := node.delegators.Get(req.GetChannel())
|
||
if !ok {
|
||
err := merr.WrapErrChannelNotFound(req.GetChannel())
|
||
log.Warn("failed to find shard cluster when sync")
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
// translate segment action
|
||
removeActions := make([]*querypb.SyncAction, 0)
|
||
group, ctx := errgroup.WithContext(ctx)
|
||
for _, action := range req.GetActions() {
|
||
log := log.With(zap.String("Action",
|
||
action.GetType().String()))
|
||
switch action.GetType() {
|
||
case querypb.SyncType_Remove:
|
||
log.Info("sync action", zap.Int64("segmentID", action.SegmentID))
|
||
removeActions = append(removeActions, action)
|
||
case querypb.SyncType_Set:
|
||
log.Info("sync action", zap.Int64("segmentID", action.SegmentID))
|
||
if action.GetInfo() == nil {
|
||
log.Warn("sync request from legacy querycoord without load info, skip")
|
||
continue
|
||
}
|
||
|
||
// to pass segment'version, we call load segment one by one
|
||
action := action
|
||
group.Go(func() error {
|
||
return shardDelegator.LoadSegments(ctx, &querypb.LoadSegmentsRequest{
|
||
Base: commonpbutil.NewMsgBase(
|
||
commonpbutil.WithMsgType(commonpb.MsgType_LoadSegments),
|
||
commonpbutil.WithMsgID(req.Base.GetMsgID()),
|
||
),
|
||
Infos: []*querypb.SegmentLoadInfo{action.GetInfo()},
|
||
Schema: req.GetSchema(),
|
||
LoadMeta: req.GetLoadMeta(),
|
||
CollectionID: req.GetCollectionID(),
|
||
ReplicaID: req.GetReplicaID(),
|
||
DstNodeID: action.GetNodeID(),
|
||
Version: action.GetVersion(),
|
||
NeedTransfer: false,
|
||
LoadScope: querypb.LoadScope_Delta,
|
||
IndexInfoList: req.GetIndexInfoList(),
|
||
})
|
||
})
|
||
case querypb.SyncType_UpdateVersion:
|
||
log.Info("sync action", zap.Int64("TargetVersion", action.GetTargetVersion()))
|
||
droppedInfos := lo.SliceToMap(action.GetDroppedInTarget(), func(id int64) (int64, uint64) {
|
||
if action.GetCheckpoint() == nil {
|
||
return id, typeutil.MaxTimestamp
|
||
}
|
||
return id, action.GetCheckpoint().Timestamp
|
||
})
|
||
shardDelegator.AddExcludedSegments(droppedInfos)
|
||
flushedInfo := lo.SliceToMap(action.GetSealedInTarget(), func(id int64) (int64, uint64) {
|
||
if action.GetCheckpoint() == nil {
|
||
return id, typeutil.MaxTimestamp
|
||
}
|
||
return id, action.GetCheckpoint().Timestamp
|
||
})
|
||
shardDelegator.AddExcludedSegments(flushedInfo)
|
||
shardDelegator.SyncTargetVersion(action.GetTargetVersion(), action.GetGrowingInTarget(),
|
||
action.GetSealedInTarget(), action.GetDroppedInTarget(), action.GetCheckpoint())
|
||
case querypb.SyncType_UpdatePartitionStats:
|
||
log.Info("sync update partition stats versions")
|
||
shardDelegator.SyncPartitionStats(ctx, action.PartitionStatsVersions)
|
||
default:
|
||
return merr.Status(merr.WrapErrServiceInternal("unknown action type", action.GetType().String())), nil
|
||
}
|
||
}
|
||
|
||
err := group.Wait()
|
||
if err != nil {
|
||
log.Warn("failed to sync distribution", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
// in case of target node offline, when try to remove segment from leader's distribution, use wildcardNodeID(-1) to skip nodeID check
|
||
for _, action := range removeActions {
|
||
shardDelegator.ReleaseSegments(ctx, &querypb.ReleaseSegmentsRequest{
|
||
NodeID: -1,
|
||
SegmentIDs: []int64{action.GetSegmentID()},
|
||
Scope: querypb.DataScope_Historical,
|
||
CollectionID: req.GetCollectionID(),
|
||
}, true)
|
||
}
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// Delete is used to forward delete message between delegator and workers.
|
||
func (node *QueryNode) Delete(ctx context.Context, req *querypb.DeleteRequest) (*commonpb.Status, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionId()),
|
||
zap.String("channel", req.GetVchannelName()),
|
||
zap.Int64("segmentID", req.GetSegmentId()),
|
||
zap.String("scope", req.GetScope().String()),
|
||
)
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return merr.Status(err), nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
log.Debug("QueryNode received worker delete detail", zap.Stringer("info", &deleteRequestStringer{DeleteRequest: req}))
|
||
|
||
filters := []segments.SegmentFilter{
|
||
segments.WithID(req.GetSegmentId()),
|
||
}
|
||
|
||
// do not add filter for Unknown & All scope, for backward cap
|
||
switch req.GetScope() {
|
||
case querypb.DataScope_Historical:
|
||
filters = append(filters, segments.WithType(segments.SegmentTypeSealed))
|
||
case querypb.DataScope_Streaming:
|
||
filters = append(filters, segments.WithType(segments.SegmentTypeGrowing))
|
||
}
|
||
|
||
segments := node.manager.Segment.GetBy(filters...)
|
||
if len(segments) == 0 {
|
||
err := merr.WrapErrSegmentNotFound(req.GetSegmentId())
|
||
log.Warn("segment not found for delete")
|
||
return merr.Status(err), nil
|
||
}
|
||
|
||
pks := storage.ParseIDs2PrimaryKeysBatch(req.GetPrimaryKeys())
|
||
for _, segment := range segments {
|
||
err := segment.Delete(ctx, pks, req.GetTimestamps())
|
||
if err != nil {
|
||
log.Warn("segment delete failed", zap.Error(err))
|
||
return merr.Status(err), nil
|
||
}
|
||
}
|
||
|
||
return merr.Success(), nil
|
||
}
|
||
|
||
// DeleteBatch is the API to apply same delete data into multiple segments.
|
||
// it's basically same as `Delete` but cost less memory pressure.
|
||
func (node *QueryNode) DeleteBatch(ctx context.Context, req *querypb.DeleteBatchRequest) (*querypb.DeleteBatchResponse, error) {
|
||
log := log.Ctx(ctx).With(
|
||
zap.Int64("collectionID", req.GetCollectionId()),
|
||
zap.String("channel", req.GetVchannelName()),
|
||
zap.Int64s("segmentIDs", req.GetSegmentIds()),
|
||
zap.String("scope", req.GetScope().String()),
|
||
)
|
||
|
||
// check node healthy
|
||
if err := node.lifetime.Add(merr.IsHealthy); err != nil {
|
||
return &querypb.DeleteBatchResponse{
|
||
Status: merr.Status(err),
|
||
}, nil
|
||
}
|
||
defer node.lifetime.Done()
|
||
|
||
// log.Debug("QueryNode received worker delete detail", zap.Stringer("info", &deleteRequestStringer{DeleteRequest: req}))
|
||
|
||
filters := []segments.SegmentFilter{
|
||
segments.WithIDs(req.GetSegmentIds()...),
|
||
}
|
||
|
||
// do not add filter for Unknown & All scope, for backward cap
|
||
switch req.GetScope() {
|
||
case querypb.DataScope_Historical:
|
||
filters = append(filters, segments.WithType(segments.SegmentTypeSealed))
|
||
case querypb.DataScope_Streaming:
|
||
filters = append(filters, segments.WithType(segments.SegmentTypeGrowing))
|
||
}
|
||
|
||
segs := node.manager.Segment.GetBy(filters...)
|
||
|
||
hitIDs := lo.Map(segs, func(segment segments.Segment, _ int) int64 {
|
||
return segment.ID()
|
||
})
|
||
// calculate missing ids, continue to delete existing ones.
|
||
missingIDs := typeutil.NewSet(req.GetSegmentIds()...).Complement(typeutil.NewSet(hitIDs...))
|
||
if missingIDs.Len() > 0 {
|
||
log.Warn("Delete batch find missing ids", zap.Int64s("missing_ids", missingIDs.Collect()))
|
||
}
|
||
|
||
pks := storage.ParseIDs2PrimaryKeysBatch(req.GetPrimaryKeys())
|
||
|
||
// control the execution batch parallel with P number
|
||
// maybe it shall be lower in case of heavy CPU usage may impacting search/query
|
||
pool := segments.GetDeletePool()
|
||
futures := make([]*conc.Future[struct{}], 0, len(segs))
|
||
errSet := typeutil.NewConcurrentSet[int64]()
|
||
|
||
for _, segment := range segs {
|
||
segment := segment
|
||
futures = append(futures, pool.Submit(func() (struct{}, error) {
|
||
// TODO @silverxia, add interface to use same data struct for segment delete
|
||
// current implementation still copys pks into protobuf(or arrow) struct
|
||
err := segment.Delete(ctx, pks, req.GetTimestamps())
|
||
if err != nil {
|
||
errSet.Insert(segment.ID())
|
||
log.Warn("segment delete failed",
|
||
zap.Int64("segmentID", segment.ID()),
|
||
zap.Error(err))
|
||
return struct{}{}, err
|
||
}
|
||
return struct{}{}, nil
|
||
}))
|
||
}
|
||
|
||
// ignore error returned, since error segment is recorded into error set
|
||
_ = conc.AwaitAll(futures...)
|
||
|
||
// return merr.Success(), nil
|
||
return &querypb.DeleteBatchResponse{
|
||
Status: merr.Success(),
|
||
FailedIds: errSet.Collect(),
|
||
}, nil
|
||
}
|
||
|
||
type deleteRequestStringer struct {
|
||
*querypb.DeleteRequest
|
||
}
|
||
|
||
func (req *deleteRequestStringer) String() string {
|
||
var pkInfo string
|
||
switch {
|
||
case req.GetPrimaryKeys().GetIntId() != nil:
|
||
ids := req.GetPrimaryKeys().GetIntId().GetData()
|
||
pkInfo = fmt.Sprintf("Pks range[%d-%d], len: %d", ids[0], ids[len(ids)-1], len(ids))
|
||
case req.GetPrimaryKeys().GetStrId() != nil:
|
||
ids := req.GetPrimaryKeys().GetStrId().GetData()
|
||
pkInfo = fmt.Sprintf("Pks range[%s-%s], len: %d", ids[0], ids[len(ids)-1], len(ids))
|
||
}
|
||
tss := req.GetTimestamps()
|
||
return fmt.Sprintf("%s, timestamp range: [%d-%d]", pkInfo, tss[0], tss[len(tss)-1])
|
||
}
|
||
|
||
func (node *QueryNode) updateDistributionModifyTS() {
|
||
node.lastModifyLock.Lock()
|
||
defer node.lastModifyLock.Unlock()
|
||
|
||
node.lastModifyTs = time.Now().UnixNano()
|
||
}
|
||
|
||
func (node *QueryNode) getDistributionModifyTS() int64 {
|
||
node.lastModifyLock.RLock()
|
||
defer node.lastModifyLock.RUnlock()
|
||
return node.lastModifyTs
|
||
}
|