milvus/internal/datanode/syncmgr/sync_manager.go
congqixia a68b32134a
fix: Verify sync task target segment and retry if not match (#30500)
See also #27675 #30469

For a sync task, the segment could be compacted during sync task. In
previous implementation, this sync task will hold only the old segment
id as KeyLock, in which case compaction on compacted to segment may run
in parallel with delta sync of this sync task.

This PR introduces sync target segment verification logic. It shall
check target segment lock it's holding beforing actually syncing logic.
If this check failed, sync task shall return`errTargetSegementNotMatch`
error and make manager re-fetch the current target segment id.

Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
2024-02-05 11:33:43 +08:00

175 lines
5.3 KiB
Go

package syncmgr
import (
"context"
"fmt"
"strconv"
"github.com/cockroachdb/errors"
"go.uber.org/zap"
"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/allocator"
"github.com/milvus-io/milvus/internal/datanode/metacache"
"github.com/milvus-io/milvus/internal/storage"
"github.com/milvus-io/milvus/pkg/config"
"github.com/milvus-io/milvus/pkg/log"
"github.com/milvus-io/milvus/pkg/util/conc"
"github.com/milvus-io/milvus/pkg/util/merr"
"github.com/milvus-io/milvus/pkg/util/paramtable"
"github.com/milvus-io/milvus/pkg/util/typeutil"
)
type SyncManagerOption struct {
chunkManager storage.ChunkManager
allocator allocator.Interface
parallelTask int
}
type SyncMeta struct {
collectionID int64
partitionID int64
segmentID int64
channelName string
schema *schemapb.CollectionSchema
checkpoint *msgpb.MsgPosition
tsFrom typeutil.Timestamp
tsTo typeutil.Timestamp
metacache metacache.MetaCache
}
// SyncMangger is the interface for sync manager.
// it processes the sync tasks inside and changes the meta.
type SyncManager interface {
// SyncData is the method to submit sync task.
SyncData(ctx context.Context, task Task) *conc.Future[error]
// GetEarliestPosition returns the earliest position (normally start position) of the processing sync task of provided channel.
GetEarliestPosition(channel string) (int64, *msgpb.MsgPosition)
// Block allows caller to block tasks of provided segment id.
// normally used by compaction task.
// if levelzero delta policy is enabled, this shall be an empty operation.
Block(segmentID int64)
// Unblock is the reverse method for `Block`.
Unblock(segmentID int64)
}
type syncManager struct {
*keyLockDispatcher[int64]
chunkManager storage.ChunkManager
allocator allocator.Interface
tasks *typeutil.ConcurrentMap[string, Task]
}
func NewSyncManager(chunkManager storage.ChunkManager, allocator allocator.Interface) (SyncManager, error) {
params := paramtable.Get()
initPoolSize := params.DataNodeCfg.MaxParallelSyncMgrTasks.GetAsInt()
if initPoolSize < 1 {
return nil, merr.WrapErrParameterInvalid("positive parallel task number", strconv.FormatInt(int64(initPoolSize), 10))
}
dispatcher := newKeyLockDispatcher[int64](initPoolSize)
log.Info("sync manager initialized", zap.Int("initPoolSize", initPoolSize))
syncMgr := &syncManager{
keyLockDispatcher: dispatcher,
chunkManager: chunkManager,
allocator: allocator,
tasks: typeutil.NewConcurrentMap[string, Task](),
}
// setup config update watcher
params.Watch(params.DataNodeCfg.MaxParallelSyncMgrTasks.Key, config.NewHandler("datanode.syncmgr.poolsize", syncMgr.resizeHandler))
return syncMgr, nil
}
func (mgr *syncManager) resizeHandler(evt *config.Event) {
if evt.HasUpdated {
log := log.Ctx(context.Background()).With(
zap.String("key", evt.Key),
zap.String("value", evt.Value),
)
size, err := strconv.ParseInt(evt.Value, 10, 64)
if err != nil {
log.Warn("failed to parse new datanode syncmgr pool size", zap.Error(err))
return
}
err = mgr.keyLockDispatcher.workerPool.Resize(int(size))
if err != nil {
log.Warn("failed to resize datanode syncmgr pool size", zap.String("key", evt.Key), zap.String("value", evt.Value), zap.Error(err))
return
}
log.Info("sync mgr pool size updated", zap.Int64("newSize", size))
}
}
func (mgr *syncManager) SyncData(ctx context.Context, task Task) *conc.Future[error] {
switch t := task.(type) {
case *SyncTask:
t.WithAllocator(mgr.allocator).WithChunkManager(mgr.chunkManager)
case *SyncTaskV2:
t.WithAllocator(mgr.allocator)
}
return mgr.safeSubmitTask(task)
}
// safeSubmitTask handles submitting task logic with optimistic target check logic
// when task returns errTargetSegmentNotMatch error
// perform refetch then retry logic
func (mgr *syncManager) safeSubmitTask(task Task) *conc.Future[error] {
taskKey := fmt.Sprintf("%d-%d", task.SegmentID(), task.Checkpoint().GetTimestamp())
mgr.tasks.Insert(taskKey, task)
return conc.Go[error](func() (error, error) {
defer mgr.tasks.Remove(taskKey)
for {
targetID, err := task.CalcTargetSegment()
if err != nil {
return err, nil
}
log.Info("task calculated target segment id",
zap.Int64("targetID", targetID),
zap.Int64("segmentID", task.SegmentID()),
)
// make sync for same segment execute in sequence
// if previous sync task is not finished, block here
f := mgr.Submit(targetID, task)
err, _ = f.Await()
if errors.Is(err, errTargetSegmentNotMatch) {
log.Info("target updated during submitting", zap.Error(err))
continue
}
return err, nil
}
})
}
func (mgr *syncManager) GetEarliestPosition(channel string) (int64, *msgpb.MsgPosition) {
var cp *msgpb.MsgPosition
var segmentID int64
mgr.tasks.Range(func(_ string, task Task) bool {
if task.StartPosition() == nil {
return true
}
if task.ChannelName() == channel {
if cp == nil || task.StartPosition().GetTimestamp() < cp.GetTimestamp() {
cp = task.StartPosition()
segmentID = task.SegmentID()
}
}
return true
})
return segmentID, cp
}
func (mgr *syncManager) Block(segmentID int64) {
mgr.keyLock.Lock(segmentID)
}
func (mgr *syncManager) Unblock(segmentID int64) {
mgr.keyLock.Unlock(segmentID)
}