github.com/whtcorpsinc/milevadb-prod@v0.0.0-20211104133533-f57f4be3b597/causetstore/milevadb-server/einsteindb/txn.go

// Copyright 2020 WHTCORPS INC, Inc.
//
// Licensed 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,
// See the License for the specific language governing permissions and
// limitations under the License.

package einsteindb

import (
	"bytes"
	"context"
	"fmt"
	"runtime/trace"
	"sort"
	"sync"
	"sync/atomic"
	"time"

	"github.com/dgryski/go-farm"
	"github.com/opentracing/opentracing-go"
	"github.com/whtcorpsinc/BerolinaSQL/terror"
	"github.com/whtcorpsinc/errors"
	"github.com/whtcorpsinc/failpoint"
	"github.com/whtcorpsinc/milevadb/ekv"
	"github.com/whtcorpsinc/milevadb/metrics"
	"github.com/whtcorpsinc/milevadb/soliton/execdetails"
	"github.com/whtcorpsinc/milevadb/soliton/logutil"
	"github.com/whtcorpsinc/milevadb/stochastikctx"
	"go.uber.org/zap"
)

var (
	_ ekv.Transaction = (*einsteindbTxn)(nil)
)

var (
	einsteindbTxnCmdHistogramWithCommit   = metrics.EinsteinDBTxnCmdHistogram.WithLabelValues(metrics.LblCommit)
	einsteindbTxnCmdHistogramWithRollback = metrics.EinsteinDBTxnCmdHistogram.WithLabelValues(metrics.LblRollback)
	einsteindbTxnCmdHistogramWithBatchGet = metrics.EinsteinDBTxnCmdHistogram.WithLabelValues(metrics.LblBatchGet)
	einsteindbTxnCmdHistogramWithGet      = metrics.EinsteinDBTxnCmdHistogram.WithLabelValues(metrics.LblGet)
)

// SchemaAmender is used by pessimistic transactions to amend commit mutations for schemaReplicant change during 2pc.
type SchemaAmender interface {
	// AmendTxn is the amend entry, new mutations will be generated based on input mutations using schemaReplicant change info.
	// The returned results are mutations need to prewrite and mutations need to cleanup.
	AmendTxn(ctx context.Context, startSchemaReplicant SchemaVer, change *RelatedSchemaChange, mutations CommitterMutations) (*CommitterMutations, error)
}

// einsteindbTxn implements ekv.Transaction.
type einsteindbTxn struct {
	snapshot    *einsteindbSnapshot
	us          ekv.UnionStore
	causetstore *einsteindbStore // for connection to region.
	startTS     uint64
	startTime   time.Time // Monotonic timestamp for recording txn time consuming.
	commitTS    uint64
	mu          sync.Mutex // For thread-safe LockKeys function.
	setCnt      int64
	vars        *ekv.Variables
	committer   *twoPhaseCommitter
	lockedCnt   int

	// For data consistency check.
	// assertions[:confirmed] is the assertion of current transaction.
	// assertions[confirmed:len(assertions)] is the assertions of current memex.
	// StmtCommit/StmtRollback may change the confirmed position.
	assertions []assertionPair
	confirmed  int

	valid bool
	dirty bool

	// txnSchemaReplicant is the schemaReplicant fetched at startTS.
	txnSchemaReplicant SchemaVer
	// SchemaAmender is used amend pessimistic txn commit mutations for schemaReplicant change
	schemaAmender SchemaAmender
	// commitCallback is called after current transaction gets committed
	commitCallback func(info ekv.TxnInfo, err error)
}

func newEinsteinDBTxn(causetstore *einsteindbStore) (*einsteindbTxn, error) {
	bo := NewBackofferWithVars(context.Background(), tsoMaxBackoff, nil)
	startTS, err := causetstore.getTimestampWithRetry(bo)
	if err != nil {
		return nil, errors.Trace(err)
	}
	return newEinsteinDBTxnWithStartTS(causetstore, startTS, causetstore.nextReplicaReadSeed())
}

// newEinsteinDBTxnWithStartTS creates a txn with startTS.
func newEinsteinDBTxnWithStartTS(causetstore *einsteindbStore, startTS uint64, replicaReadSeed uint32) (*einsteindbTxn, error) {
	ver := ekv.NewVersion(startTS)
	snapshot := newEinsteinDBSnapshot(causetstore, ver, replicaReadSeed)
	return &einsteindbTxn{
		snapshot:    snapshot,
		us:          ekv.NewUnionStore(snapshot),
		causetstore: causetstore,
		startTS:     startTS,
		startTime:   time.Now(),
		valid:       true,
		vars:        ekv.DefaultVars,
	}, nil
}

type assertionPair struct {
	key       ekv.Key
	assertion ekv.AssertionType
}

func (a assertionPair) String() string {
	return fmt.Sprintf("key: %s, assertion type: %d", a.key, a.assertion)
}

// SetSuccess is used to probe if ekv variables are set or not. It is ONLY used in test cases.
var SetSuccess = false

func (txn *einsteindbTxn) SetVars(vars *ekv.Variables) {
	txn.vars = vars
	txn.snapshot.vars = vars
	failpoint.Inject("probeSetVars", func(val failpoint.Value) {
		if val.(bool) {
			SetSuccess = true
		}
	})
}

func (txn *einsteindbTxn) GetVars() *ekv.Variables {
	return txn.vars
}

// Get implements transaction interface.
func (txn *einsteindbTxn) Get(ctx context.Context, k ekv.Key) ([]byte, error) {
	ret, err := txn.us.Get(ctx, k)
	if ekv.IsErrNotFound(err) {
		return nil, err
	}
	if err != nil {
		return nil, errors.Trace(err)
	}

	return ret, nil
}

func (txn *einsteindbTxn) BatchGet(ctx context.Context, keys []ekv.Key) (map[string][]byte, error) {
	if span := opentracing.SpanFromContext(ctx); span != nil && span.Tracer() != nil {
		span1 := span.Tracer().StartSpan("einsteindbTxn.BatchGet", opentracing.ChildOf(span.Context()))
		defer span1.Finish()
		ctx = opentracing.ContextWithSpan(ctx, span1)
	}
	return ekv.NewBufferBatchGetter(txn.GetMemBuffer(), nil, txn.snapshot).BatchGet(ctx, keys)
}

func (txn *einsteindbTxn) Set(k ekv.Key, v []byte) error {
	txn.setCnt++
	return txn.us.GetMemBuffer().Set(k, v)
}

func (txn *einsteindbTxn) String() string {
	return fmt.Sprintf("%d", txn.StartTS())
}

func (txn *einsteindbTxn) Iter(k ekv.Key, upperBound ekv.Key) (ekv.Iterator, error) {
	return txn.us.Iter(k, upperBound)
}

// IterReverse creates a reversed Iterator positioned on the first entry which key is less than k.
func (txn *einsteindbTxn) IterReverse(k ekv.Key) (ekv.Iterator, error) {
	return txn.us.IterReverse(k)
}

func (txn *einsteindbTxn) Delete(k ekv.Key) error {
	return txn.us.GetMemBuffer().Delete(k)
}

func (txn *einsteindbTxn) SetOption(opt ekv.Option, val interface{}) {
	txn.us.SetOption(opt, val)
	txn.snapshot.SetOption(opt, val)
	switch opt {
	case ekv.SchemaReplicant:
		txn.txnSchemaReplicant = val.(SchemaVer)
	case ekv.SchemaAmender:
		txn.schemaAmender = val.(SchemaAmender)
	case ekv.CommitHook:
		txn.commitCallback = val.(func(info ekv.TxnInfo, err error))
	}
}

func (txn *einsteindbTxn) DelOption(opt ekv.Option) {
	txn.us.DelOption(opt)
}

func (txn *einsteindbTxn) IsPessimistic() bool {
	return txn.us.GetOption(ekv.Pessimistic) != nil
}

func (txn *einsteindbTxn) Commit(ctx context.Context) error {
	if span := opentracing.SpanFromContext(ctx); span != nil && span.Tracer() != nil {
		span1 := span.Tracer().StartSpan("einsteindbTxn.Commit", opentracing.ChildOf(span.Context()))
		defer span1.Finish()
		ctx = opentracing.ContextWithSpan(ctx, span1)
	}
	defer trace.StartRegion(ctx, "CommitTxn").End()

	if !txn.valid {
		return ekv.ErrInvalidTxn
	}
	defer txn.close()

	failpoint.Inject("mockCommitError", func(val failpoint.Value) {
		if val.(bool) && ekv.IsMockCommitErrorEnable() {
			ekv.MockCommitErrorDisable()
			failpoint.Return(errors.New("mock commit error"))
		}
	})

	start := time.Now()
	defer func() { einsteindbTxnCmdHistogramWithCommit.Observe(time.Since(start).Seconds()) }()

	// connID is used for log.
	var connID uint64
	val := ctx.Value(stochastikctx.ConnID)
	if val != nil {
		connID = val.(uint64)
	}

	var err error
	// If the txn use pessimistic dagger, committer is initialized.
	committer := txn.committer
	if committer == nil {
		committer, err = newTwoPhaseCommitter(txn, connID)
		if err != nil {
			return errors.Trace(err)
		}
	}
	defer func() {
		// For async commit transactions, the ttl manager will be closed in the asynchronous commit goroutine.
		if !committer.isAsyncCommit() {
			committer.ttlManager.close()
		}
	}()

	initRegion := trace.StartRegion(ctx, "InitKeys")
	err = committer.initKeysAndMutations()
	initRegion.End()
	if err != nil {
		return errors.Trace(err)
	}
	if committer.mutations.len() == 0 {
		return nil
	}

	defer func() {
		ctxValue := ctx.Value(execdetails.CommitDetailCtxKey)
		if ctxValue != nil {
			commitDetail := ctxValue.(**execdetails.CommitDetails)
			if *commitDetail != nil {
				(*commitDetail).TxnRetry++
			} else {
				*commitDetail = committer.getDetail()
			}
		}
	}()
	// latches disabled
	// pessimistic transaction should also bypass latch.
	if txn.causetstore.txnLatches == nil || txn.IsPessimistic() {
		err = committer.execute(ctx)
		if val == nil || connID > 0 {
			txn.onCommitted(err)
		}
		logutil.Logger(ctx).Debug("[ekv] txnLatches disabled, 2pc directly", zap.Error(err))
		return errors.Trace(err)
	}

	// latches enabled
	// for transactions which need to acquire latches
	start = time.Now()
	dagger := txn.causetstore.txnLatches.Lock(committer.startTS, committer.mutations.keys)
	commitDetail := committer.getDetail()
	commitDetail.LocalLatchTime = time.Since(start)
	if commitDetail.LocalLatchTime > 0 {
		metrics.EinsteinDBLocalLatchWaitTimeHistogram.Observe(commitDetail.LocalLatchTime.Seconds())
	}
	defer txn.causetstore.txnLatches.UnLock(dagger)
	if dagger.IsStale() {
		return ekv.ErrWriteConflictInMilevaDB.FastGenByArgs(txn.startTS)
	}
	err = committer.execute(ctx)
	if val == nil || connID > 0 {
		txn.onCommitted(err)
	}
	if err == nil {
		dagger.SetCommitTS(committer.commitTS)
	}
	logutil.Logger(ctx).Debug("[ekv] txnLatches enabled while txn retryable", zap.Error(err))
	return errors.Trace(err)
}

func (txn *einsteindbTxn) close() {
	txn.valid = false
}

func (txn *einsteindbTxn) Rollback() error {
	if !txn.valid {
		return ekv.ErrInvalidTxn
	}
	start := time.Now()
	// Clean up pessimistic dagger.
	if txn.IsPessimistic() && txn.committer != nil {
		err := txn.rollbackPessimisticLocks()
		txn.committer.ttlManager.close()
		if err != nil {
			logutil.BgLogger().Error(err.Error())
		}
	}
	txn.close()
	logutil.BgLogger().Debug("[ekv] rollback txn", zap.Uint64("txnStartTS", txn.StartTS()))
	einsteindbTxnCmdHistogramWithRollback.Observe(time.Since(start).Seconds())
	return nil
}

func (txn *einsteindbTxn) rollbackPessimisticLocks() error {
	if txn.lockedCnt == 0 {
		return nil
	}
	bo := NewBackofferWithVars(context.Background(), cleanupMaxBackoff, txn.vars)
	keys := txn.collectLockedKeys()
	return txn.committer.pessimisticRollbackMutations(bo, CommitterMutations{keys: keys})
}

func (txn *einsteindbTxn) collectLockedKeys() [][]byte {
	keys := make([][]byte, 0, txn.lockedCnt)
	buf := txn.GetMemBuffer()
	var err error
	for it := buf.IterWithFlags(nil, nil); it.Valid(); err = it.Next() {
		_ = err
		if it.Flags().HasLocked() {
			keys = append(keys, it.Key())
		}
	}
	return keys
}

func (txn *einsteindbTxn) onCommitted(err error) {
	if txn.commitCallback != nil {
		info := ekv.TxnInfo{StartTS: txn.startTS, CommitTS: txn.commitTS}
		if err != nil {
			info.ErrMsg = err.Error()
		}
		txn.commitCallback(info, err)
	}
}

// lockWaitTime in ms, except that ekv.LockAlwaysWait(0) means always wait dagger, ekv.LockNowait(-1) means nowait dagger
func (txn *einsteindbTxn) LockKeys(ctx context.Context, lockCtx *ekv.LockCtx, keysInput ...ekv.Key) error {
	// Exclude keys that are already locked.
	var err error
	keys := make([][]byte, 0, len(keysInput))
	startTime := time.Now()
	txn.mu.Lock()
	defer txn.mu.Unlock()
	defer func() {
		if err == nil {
			if lockCtx.PessimisticLockWaited != nil {
				if atomic.LoadInt32(lockCtx.PessimisticLockWaited) > 0 {
					timeWaited := time.Since(lockCtx.WaitStartTime)
					atomic.StoreInt64(lockCtx.LockKeysDuration, int64(timeWaited))
					metrics.EinsteinDBPessimisticLockKeysDuration.Observe(timeWaited.Seconds())
				}
			}
		}
		if lockCtx.LockKeysCount != nil {
			*lockCtx.LockKeysCount += int32(len(keys))
		}
		if lockCtx.Stats != nil {
			lockCtx.Stats.TotalTime = time.Since(startTime)
			ctxValue := ctx.Value(execdetails.LockKeysDetailCtxKey)
			if ctxValue != nil {
				lockKeysDetail := ctxValue.(**execdetails.LockKeysDetails)
				*lockKeysDetail = lockCtx.Stats
			}
		}
	}()
	memBuf := txn.us.GetMemBuffer()
	for _, key := range keysInput {
		// The value of lockedMap is only used by pessimistic transactions.
		var valueExist, locked, checkKeyExists bool
		if flags, err := memBuf.GetFlags(key); err == nil {
			locked = flags.HasLocked()
			valueExist = flags.HasLockedValueExists()
			checkKeyExists = flags.HasNeedCheckExists()
		}
		if !locked {
			keys = append(keys, key)
		} else if txn.IsPessimistic() {
			if checkKeyExists && valueExist {
				return txn.committer.extractKeyExistsErr(key)
			}
		}
		if lockCtx.ReturnValues && locked {
			// An already locked key can not return values, we add an entry to let the caller get the value
			// in other ways.
			lockCtx.Values[string(key)] = ekv.ReturnedValue{AlreadyLocked: true}
		}
	}
	if len(keys) == 0 {
		return nil
	}
	keys = deduplicateKeys(keys)
	if txn.IsPessimistic() && lockCtx.ForUFIDelateTS > 0 {
		if txn.committer == nil {
			// connID is used for log.
			var connID uint64
			var err error
			val := ctx.Value(stochastikctx.ConnID)
			if val != nil {
				connID = val.(uint64)
			}
			txn.committer, err = newTwoPhaseCommitter(txn, connID)
			if err != nil {
				return err
			}
		}
		var assignedPrimaryKey bool
		if txn.committer.primaryKey == nil {
			txn.committer.primaryKey = keys[0]
			assignedPrimaryKey = true
		}

		lockCtx.Stats = &execdetails.LockKeysDetails{
			LockKeys: int32(len(keys)),
		}
		bo := NewBackofferWithVars(ctx, pessimisticLockMaxBackoff, txn.vars)
		txn.committer.forUFIDelateTS = lockCtx.ForUFIDelateTS
		// If the number of keys greater than 1, it can be on different region,
		// concurrently execute on multiple regions may lead to deadlock.
		txn.committer.isFirstLock = txn.lockedCnt == 0 && len(keys) == 1
		err = txn.committer.pessimisticLockMutations(bo, lockCtx, CommitterMutations{keys: keys})
		if bo.totalSleep > 0 {
			atomic.AddInt64(&lockCtx.Stats.BackoffTime, int64(bo.totalSleep)*int64(time.Millisecond))
			lockCtx.Stats.Mu.Lock()
			lockCtx.Stats.Mu.BackoffTypes = append(lockCtx.Stats.Mu.BackoffTypes, bo.types...)
			lockCtx.Stats.Mu.Unlock()
		}
		if lockCtx.Killed != nil {
			// If the kill signal is received during waiting for pessimisticLock,
			// pessimisticLockKeys would handle the error but it doesn't reset the flag.
			// We need to reset the killed flag here.
			atomic.CompareAndSwapUint32(lockCtx.Killed, 1, 0)
		}
		if err != nil {
			for _, key := range keys {
				if txn.us.HasPresumeKeyNotExists(key) {
					txn.us.UnmarkPresumeKeyNotExists(key)
				}
			}
			keyMayBeLocked := terror.ErrorNotEqual(ekv.ErrWriteConflict, err) && terror.ErrorNotEqual(ekv.ErrKeyExists, err)
			// If there is only 1 key and dagger fails, no need to do pessimistic rollback.
			if len(keys) > 1 || keyMayBeLocked {
				wg := txn.asyncPessimisticRollback(ctx, keys)
				if dl, ok := errors.Cause(err).(*ErrDeadlock); ok && hashInKeys(dl.DeadlockKeyHash, keys) {
					dl.IsRetryable = true
					// Wait for the pessimistic rollback to finish before we retry the memex.
					wg.Wait()
					// Sleep a little, wait for the other transaction that blocked by this transaction to acquire the dagger.
					time.Sleep(time.Millisecond * 5)
					failpoint.Inject("SingleStmtDeadLockRetrySleep", func() {
						time.Sleep(300 * time.Millisecond)
					})
				}
			}
			if assignedPrimaryKey {
				// unset the primary key if we assigned primary key when failed to dagger it.
				txn.committer.primaryKey = nil
			}
			return err
		}
		if assignedPrimaryKey {
			txn.committer.ttlManager.run(txn.committer, lockCtx)
		}
	}
	for _, key := range keys {
		valExists := ekv.SetKeyLockedValueExists
		// PointGet and BatchPointGet will return value in pessimistic dagger response, the value may not exist.
		// For other dagger modes, the locked key values always exist.
		if lockCtx.ReturnValues {
			val, _ := lockCtx.Values[string(key)]
			if len(val.Value) == 0 {
				valExists = ekv.SetKeyLockedValueNotExists
			}
		}
		memBuf.UFIDelateFlags(key, ekv.SetKeyLocked, ekv.DelNeedCheckExists, valExists)
	}
	txn.lockedCnt += len(keys)
	return nil
}

// deduplicateKeys deduplicate the keys, it use sort instead of map to avoid memory allocation.
func deduplicateKeys(keys [][]byte) [][]byte {
	sort.Slice(keys, func(i, j int) bool {
		return bytes.Compare(keys[i], keys[j]) < 0
	})
	deduped := keys[:1]
	for i := 1; i < len(keys); i++ {
		if !bytes.Equal(deduped[len(deduped)-1], keys[i]) {
			deduped = append(deduped, keys[i])
		}
	}
	return deduped
}

func (txn *einsteindbTxn) asyncPessimisticRollback(ctx context.Context, keys [][]byte) *sync.WaitGroup {
	// Clone a new committer for execute in background.
	committer := &twoPhaseCommitter{
		causetstore:    txn.committer.causetstore,
		connID:         txn.committer.connID,
		startTS:        txn.committer.startTS,
		forUFIDelateTS: txn.committer.forUFIDelateTS,
		primaryKey:     txn.committer.primaryKey,
	}
	wg := new(sync.WaitGroup)
	wg.Add(1)
	go func() {
		failpoint.Inject("AsyncRollBackSleep", func() {
			time.Sleep(100 * time.Millisecond)
		})
		err := committer.pessimisticRollbackMutations(NewBackofferWithVars(ctx, pessimisticRollbackMaxBackoff, txn.vars), CommitterMutations{keys: keys})
		if err != nil {
			logutil.Logger(ctx).Warn("[ekv] pessimisticRollback failed.", zap.Error(err))
		}
		wg.Done()
	}()
	return wg
}

func hashInKeys(deadlockKeyHash uint64, keys [][]byte) bool {
	for _, key := range keys {
		if farm.Fingerprint64(key) == deadlockKeyHash {
			return true
		}
	}
	return false
}

func (txn *einsteindbTxn) IsReadOnly() bool {
	return !txn.us.GetMemBuffer().Dirty()
}

func (txn *einsteindbTxn) StartTS() uint64 {
	return txn.startTS
}

func (txn *einsteindbTxn) Valid() bool {
	return txn.valid
}

func (txn *einsteindbTxn) Len() int {
	return txn.us.GetMemBuffer().Len()
}

func (txn *einsteindbTxn) Size() int {
	return txn.us.GetMemBuffer().Size()
}

func (txn *einsteindbTxn) Reset() {
	txn.us.GetMemBuffer().Reset()
}

func (txn *einsteindbTxn) GetUnionStore() ekv.UnionStore {
	return txn.us
}

func (txn *einsteindbTxn) GetMemBuffer() ekv.MemBuffer {
	return txn.us.GetMemBuffer()
}

func (txn *einsteindbTxn) GetSnapshot() ekv.Snapshot {
	return txn.snapshot
}