reload region cache when store is resolved from invalid status (tikv#843

)

Signed-off-by: you06 <you1474600@gmail.com>
Co-authored-by: disksing <i@disksing.com>

error/error.go

@@ -246,7 +246,7 @@ type ErrAssertionFailed struct {
 	*kvrpcpb.AssertionFailed
 }
 
-// ErrLockOnlyIfExistsNoReturnValue is used when the flag `LockOnlyIfExists` of `LockCtx` is set, but `ReturnValues“ is not.
+// ErrLockOnlyIfExistsNoReturnValue is used when the flag `LockOnlyIfExists` of `LockCtx` is set, but `ReturnValues` is not.
 type ErrLockOnlyIfExistsNoReturnValue struct {
 	StartTS     uint64
 	ForUpdateTs uint64

internal/locate/region_cache.go

@@ -153,6 +153,7 @@ type Region struct {
 	syncFlag      int32          // region need be sync in next turn
 	lastAccess    int64          // last region access time, see checkRegionCacheTTL
 	invalidReason InvalidReason  // the reason why the region is invalidated
+	asyncReload   atomic.Bool    // the region need to be reloaded in async mode
 }
 
 // AccessIndex represent the index for accessIndex array
@@ -411,7 +412,7 @@ func newRegionIndexMu(rs []*Region) *regionIndexMu {
 	r.latestVersions = make(map[uint64]RegionVerID)
 	r.sorted = NewSortedRegions(btreeDegree)
 	for _, region := range rs {
-		r.insertRegionToCache(region)
+		r.insertRegionToCache(region, true)
 	}
 	return r
 }
@@ -457,6 +458,11 @@ type RegionCache struct {
 		// requestLiveness always returns unreachable.
 		mockRequestLiveness atomic.Pointer[livenessFunc]
 	}
+
+	regionsNeedReload struct {
+		sync.Mutex
+		regions []uint64
+	}
 }
 
 // NewRegionCache creates a RegionCache.
@@ -510,8 +516,8 @@ func (c *RegionCache) clear() {
 }
 
 // thread unsafe, should use with lock
-func (c *RegionCache) insertRegionToCache(cachedRegion *Region) {
-	c.mu.insertRegionToCache(cachedRegion)
+func (c *RegionCache) insertRegionToCache(cachedRegion *Region, invalidateOldRegion bool) {
+	c.mu.insertRegionToCache(cachedRegion, invalidateOldRegion)
 }
 
 // Close releases region cache's resource.
@@ -522,8 +528,13 @@ func (c *RegionCache) Close() {
 // asyncCheckAndResolveLoop with
 func (c *RegionCache) asyncCheckAndResolveLoop(interval time.Duration) {
 	ticker := time.NewTicker(interval)
-	defer ticker.Stop()
+	reloadRegionTicker := time.NewTicker(10 * time.Second)
+	defer func() {
+		ticker.Stop()
+		reloadRegionTicker.Stop()
+	}()
 	var needCheckStores []*Store
+	reloadNextLoop := make(map[uint64]struct{})
 	for {
 		needCheckStores = needCheckStores[:0]
 		select {
@@ -541,6 +552,21 @@ func (c *RegionCache) asyncCheckAndResolveLoop(interval time.Duration) {
 				// there's a deleted store in the stores map which guaranteed by reReslve().
 				return state != unresolved && state != tombstone && state != deleted
 			})
+		case <-reloadRegionTicker.C:
+			for regionID := range reloadNextLoop {
+				c.reloadRegion(regionID)
+				delete(reloadNextLoop, regionID)
+			}
+			c.regionsNeedReload.Lock()
+			for _, regionID := range c.regionsNeedReload.regions {
+				// will reload in next tick, wait a while for two reasons:
+				// 1. there may an unavailable duration while recreating the connection.
+				// 2. the store may just be started, and wait safe ts synced to avoid the
+				// possible dataIsNotReady error.
+				reloadNextLoop[regionID] = struct{}{}
+			}
+			c.regionsNeedReload.regions = c.regionsNeedReload.regions[:0]
+			c.regionsNeedReload.Unlock()
 		}
 	}
 }
@@ -1043,7 +1069,7 @@ func (c *RegionCache) findRegionByKey(bo *retry.Backoffer, key []byte, isEndKey
 		logutil.Eventf(bo.GetCtx(), "load region %d from pd, due to cache-miss", lr.GetID())
 		r = lr
 		c.mu.Lock()
-		c.insertRegionToCache(r)
+		c.insertRegionToCache(r, true)
 		c.mu.Unlock()
 	} else if r.checkNeedReloadAndMarkUpdated() {
 		// load region when it be marked as need reload.
@@ -1056,7 +1082,7 @@ func (c *RegionCache) findRegionByKey(bo *retry.Backoffer, key []byte, isEndKey
 			logutil.Eventf(bo.GetCtx(), "load region %d from pd, due to need-reload", lr.GetID())
 			r = lr
 			c.mu.Lock()
-			c.insertRegionToCache(r)
+			c.insertRegionToCache(r, true)
 			c.mu.Unlock()
 		}
 	}
@@ -1197,7 +1223,7 @@ func (c *RegionCache) LocateRegionByID(bo *retry.Backoffer, regionID uint64) (*K
 			} else {
 				r = lr
 				c.mu.Lock()
-				c.insertRegionToCache(r)
+				c.insertRegionToCache(r, true)
 				c.mu.Unlock()
 			}
 		}
@@ -1216,7 +1242,7 @@ func (c *RegionCache) LocateRegionByID(bo *retry.Backoffer, regionID uint64) (*K
 	}
 
 	c.mu.Lock()
-	c.insertRegionToCache(r)
+	c.insertRegionToCache(r, true)
 	c.mu.Unlock()
 	return &KeyLocation{
 		Region:   r.VerID(),
@@ -1226,6 +1252,36 @@ func (c *RegionCache) LocateRegionByID(bo *retry.Backoffer, regionID uint64) (*K
 	}, nil
 }
 
+func (c *RegionCache) scheduleReloadRegion(region *Region) {
+	if region == nil || !region.asyncReload.CompareAndSwap(false, true) {
+		// async reload scheduled by other thread.
+		return
+	}
+	regionID := region.GetID()
+	if regionID > 0 {
+		c.regionsNeedReload.Lock()
+		c.regionsNeedReload.regions = append(c.regionsNeedReload.regions, regionID)
+		c.regionsNeedReload.Unlock()
+	}
+}
+
+func (c *RegionCache) reloadRegion(regionID uint64) {
+	bo := retry.NewNoopBackoff(context.Background())
+	lr, err := c.loadRegionByID(bo, regionID)
+	if err != nil {
+		// ignore error and use old region info.
+		logutil.Logger(bo.GetCtx()).Error("load region failure",
+			zap.Uint64("regionID", regionID), zap.Error(err))
+		if oldRegion := c.getRegionByIDFromCache(regionID); oldRegion != nil {
+			oldRegion.asyncReload.Store(false)
+		}
+		return
+	}
+	c.mu.Lock()
+	c.insertRegionToCache(lr, false)
+	c.mu.Unlock()
+}
+
 // GroupKeysByRegion separates keys into groups by their belonging Regions.
 // Specially it also returns the first key's region which may be used as the
 // 'PrimaryLockKey' and should be committed ahead of others.
@@ -1310,7 +1366,7 @@ func (c *RegionCache) BatchLoadRegionsWithKeyRange(bo *retry.Backoffer, startKey
 	// TODO(youjiali1995): scanRegions always fetch regions from PD and these regions don't contain buckets information
 	// for less traffic, so newly inserted regions in region cache don't have buckets information. We should improve it.
 	for _, region := range regions {
-		c.insertRegionToCache(region)
+		c.insertRegionToCache(region, true)
 	}
 
 	return
@@ -1384,7 +1440,9 @@ func (mu *regionIndexMu) removeVersionFromCache(oldVer RegionVerID, regionID uin
 
 // insertRegionToCache tries to insert the Region to cache.
 // It should be protected by c.mu.l.Lock().
-func (mu *regionIndexMu) insertRegionToCache(cachedRegion *Region) {
+// if `invalidateOldRegion` is false, the old region cache should be still valid,
+// and it may still be used by some kv requests.
+func (mu *regionIndexMu) insertRegionToCache(cachedRegion *Region, invalidateOldRegion bool) {
 	oldRegion := mu.sorted.ReplaceOrInsert(cachedRegion)
 	if oldRegion != nil {
 		store := cachedRegion.getStore()
@@ -1399,8 +1457,11 @@ func (mu *regionIndexMu) insertRegionToCache(cachedRegion *Region) {
 		if InvalidReason(atomic.LoadInt32((*int32)(&oldRegion.invalidReason))) == NoLeader {
 			store.workTiKVIdx = (oldRegionStore.workTiKVIdx + 1) % AccessIndex(store.accessStoreNum(tiKVOnly))
 		}
-		// Invalidate the old region in case it's not invalidated and some requests try with the stale region information.
-		oldRegion.invalidate(Other)
+		// If the old region is still valid, do not invalidate it to avoid unnecessary backoff.
+		if invalidateOldRegion {
+			// Invalidate the old region in case it's not invalidated and some requests try with the stale region information.
+			oldRegion.invalidate(Other)
+		}
 		// Don't refresh TiFlash work idx for region. Otherwise, it will always goto a invalid store which
 		// is under transferring regions.
 		store.workTiFlashIdx.Store(oldRegionStore.workTiFlashIdx.Load())
@@ -1919,7 +1980,7 @@ func (c *RegionCache) OnRegionEpochNotMatch(bo *retry.Backoffer, ctx *RPCContext
 
 	c.mu.Lock()
 	for _, region := range newRegions {
-		c.insertRegionToCache(region)
+		c.insertRegionToCache(region, true)
 	}
 	c.mu.Unlock()
 
@@ -2037,7 +2098,7 @@ func (c *RegionCache) UpdateBucketsIfNeeded(regionID RegionVerID, latestBucketsV
 				return
 			}
 			c.mu.Lock()
-			c.insertRegionToCache(new)
+			c.insertRegionToCache(new, true)
 			c.mu.Unlock()
 		}()
 	}
@@ -2525,8 +2586,8 @@ func (s *Store) reResolve(c *RegionCache) (bool, error) {
 }
 
 func (s *Store) getResolveState() resolveState {
-	var state resolveState
 	if s == nil {
+		var state resolveState
 		return state
 	}
 	return resolveState(atomic.LoadUint64(&s.state))

internal/locate/region_cache_test.go

@@ -966,7 +966,7 @@ func (s *testRegionCacheSuite) TestRegionEpochAheadOfTiKV() {
 	region := createSampleRegion([]byte("k1"), []byte("k2"))
 	region.meta.Id = 1
 	region.meta.RegionEpoch = &metapb.RegionEpoch{Version: 10, ConfVer: 10}
-	cache.insertRegionToCache(region)
+	cache.insertRegionToCache(region, true)
 
 	r1 := metapb.Region{Id: 1, RegionEpoch: &metapb.RegionEpoch{Version: 9, ConfVer: 10}}
 	r2 := metapb.Region{Id: 1, RegionEpoch: &metapb.RegionEpoch{Version: 10, ConfVer: 9}}
@@ -1257,7 +1257,7 @@ func (s *testRegionCacheSuite) TestPeersLenChange() {
 	filterUnavailablePeers(cpRegion)
 	region, err := newRegion(s.bo, s.cache, cpRegion)
 	s.Nil(err)
-	s.cache.insertRegionToCache(region)
+	s.cache.insertRegionToCache(region, true)
 
 	// OnSendFail should not panic
 	s.cache.OnSendFail(retry.NewNoopBackoff(context.Background()), ctx, false, errors.New("send fail"))
@@ -1293,7 +1293,7 @@ func (s *testRegionCacheSuite) TestPeersLenChangedByWitness() {
 	cpRegion := &pd.Region{Meta: cpMeta}
 	region, err := newRegion(s.bo, s.cache, cpRegion)
 	s.Nil(err)
-	s.cache.insertRegionToCache(region)
+	s.cache.insertRegionToCache(region, true)
 
 	// OnSendFail should not panic
 	s.cache.OnSendFail(retry.NewNoopBackoff(context.Background()), ctx, false, errors.New("send fail"))
@@ -1466,12 +1466,12 @@ func (s *testRegionCacheSuite) TestBuckets() {
 	fakeRegion.setStore(cachedRegion.getStore().clone())
 	// no buckets
 	fakeRegion.getStore().buckets = nil
-	s.cache.insertRegionToCache(fakeRegion)
+	s.cache.insertRegionToCache(fakeRegion, true)
 	cachedRegion = s.getRegion([]byte("a"))
 	s.Equal(defaultBuckets, cachedRegion.getStore().buckets)
 	// stale buckets
 	fakeRegion.getStore().buckets = &metapb.Buckets{Version: defaultBuckets.Version - 1}
-	s.cache.insertRegionToCache(fakeRegion)
+	s.cache.insertRegionToCache(fakeRegion, true)
 	cachedRegion = s.getRegion([]byte("a"))
 	s.Equal(defaultBuckets, cachedRegion.getStore().buckets)
 	// new buckets
@@ -1481,7 +1481,7 @@ func (s *testRegionCacheSuite) TestBuckets() {
 		Keys:     buckets.Keys,
 	}
 	fakeRegion.getStore().buckets = newBuckets
-	s.cache.insertRegionToCache(fakeRegion)
+	s.cache.insertRegionToCache(fakeRegion, true)
 	cachedRegion = s.getRegion([]byte("a"))
 	s.Equal(newBuckets, cachedRegion.getStore().buckets)
 
@@ -1614,7 +1614,7 @@ func (s *testRegionCacheSuite) TestRemoveIntersectingRegions() {
 	region, err := s.cache.loadRegion(s.bo, []byte("c"), false)
 	s.Nil(err)
 	s.Equal(region.GetID(), regions[0])
-	s.cache.insertRegionToCache(region)
+	s.cache.insertRegionToCache(region, true)
 	loc, err = s.cache.LocateKey(s.bo, []byte{'c'})
 	s.Nil(err)
 	s.Equal(loc.Region.GetID(), regions[0])
@@ -1625,7 +1625,7 @@ func (s *testRegionCacheSuite) TestRemoveIntersectingRegions() {
 	region, err = s.cache.loadRegion(s.bo, []byte("e"), false)
 	s.Nil(err)
 	s.Equal(region.GetID(), regions[0])
-	s.cache.insertRegionToCache(region)
+	s.cache.insertRegionToCache(region, true)
 	loc, err = s.cache.LocateKey(s.bo, []byte{'e'})
 	s.Nil(err)
 	s.Equal(loc.Region.GetID(), regions[0])
@@ -1739,7 +1739,7 @@ func (s *testRegionRequestToSingleStoreSuite) TestRefreshCache() {
 	v2 := region.Region.confVer + 1
 	r2 := metapb.Region{Id: region.Region.id, RegionEpoch: &metapb.RegionEpoch{Version: region.Region.ver, ConfVer: v2}, StartKey: []byte{1}}
 	st := &Store{storeID: s.store}
-	s.cache.insertRegionToCache(&Region{meta: &r2, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()})
+	s.cache.insertRegionToCache(&Region{meta: &r2, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()}, true)
 
 	r, _ = s.cache.scanRegionsFromCache(s.bo, []byte{}, nil, 10)
 	s.Equal(len(r), 2)

internal/locate/region_request.go

@@ -573,18 +573,39 @@ func (state *accessFollower) next(bo *retry.Backoffer, selector *replicaSelector
 	if state.option.preferLeader {
 		state.lastIdx = state.leaderIdx
 	}
+	var offset int
+	if state.lastIdx >= 0 {
+		offset = rand.Intn(replicaSize)
+	}
+	reloadRegion := false
 	for i := 0; i < replicaSize && !state.option.leaderOnly; i++ {
-		idx := AccessIndex((int(state.lastIdx) + i) % replicaSize)
-		// If the given store is abnormal to be accessed under `ReplicaReadMixed` mode, we should choose other followers or leader
-		// as candidates to serve the Read request. Meanwhile, we should make the choice of next() meet Uniform Distribution.
-		for cnt := 0; cnt < replicaSize && !state.isCandidate(idx, selector.replicas[idx]); cnt++ {
-			idx = AccessIndex((int(idx) + rand.Intn(replicaSize)) % replicaSize)
+		var idx AccessIndex
+		if state.option.preferLeader {
+			if i == 0 {
+				idx = state.lastIdx
+			} else {
+				// randomly select next replica, but skip state.lastIdx
+				if (i+offset)%replicaSize == 0 {
+					offset++
+				}
+			}
+		} else {
+			idx = AccessIndex((int(state.lastIdx) + i) % replicaSize)
 		}
-		if state.isCandidate(idx, selector.replicas[idx]) {
+		selectReplica := selector.replicas[idx]
+		if state.isCandidate(idx, selectReplica) {
 			state.lastIdx = idx
 			selector.targetIdx = idx
 			break
 		}
+		if selectReplica.isEpochStale() &&
+			selectReplica.store.getResolveState() == resolved &&
+			selectReplica.store.getLivenessState() == reachable {
+			reloadRegion = true
+		}
+	}
+	if reloadRegion {
+		selector.regionCache.scheduleReloadRegion(selector.region)
 	}
 	// If there is no candidate, fallback to the leader.
 	if selector.targetIdx < 0 {

internal/locate/region_request3_test.go

@@ -328,7 +328,7 @@ func (s *testRegionRequestToThreeStoresSuite) TestLearnerReplicaSelector() {
 	cache := NewRegionCache(s.cache.pdClient)
 	defer cache.Close()
 	cache.mu.Lock()
-	cache.insertRegionToCache(region)
+	cache.insertRegionToCache(region, true)
 	cache.mu.Unlock()
 
 	// Test accessFollower state with kv.ReplicaReadLearner request type.
@@ -379,7 +379,7 @@ func (s *testRegionRequestToThreeStoresSuite) TestReplicaSelector() {
 	cache := NewRegionCache(s.cache.pdClient)
 	defer cache.Close()
 	cache.mu.Lock()
-	cache.insertRegionToCache(region)
+	cache.insertRegionToCache(region, true)
 	cache.mu.Unlock()
 
 	// Verify creating the replicaSelector.

internal/locate/region_request_test.go

@@ -608,7 +608,7 @@ func (s *testRegionRequestToSingleStoreSuite) TestGetRegionByIDFromCache() {
 	v2 := region.Region.confVer + 1
 	r2 := metapb.Region{Id: region.Region.id, RegionEpoch: &metapb.RegionEpoch{Version: region.Region.ver, ConfVer: v2}, StartKey: []byte{1}}
 	st := &Store{storeID: s.store}
-	s.cache.insertRegionToCache(&Region{meta: &r2, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()})
+	s.cache.insertRegionToCache(&Region{meta: &r2, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()}, true)
 	region, err = s.cache.LocateRegionByID(s.bo, s.region)
 	s.Nil(err)
 	s.NotNil(region)
@@ -618,7 +618,7 @@ func (s *testRegionRequestToSingleStoreSuite) TestGetRegionByIDFromCache() {
 	v3 := region.Region.confVer + 1
 	r3 := metapb.Region{Id: region.Region.id, RegionEpoch: &metapb.RegionEpoch{Version: v3, ConfVer: region.Region.confVer}, StartKey: []byte{2}}
 	st = &Store{storeID: s.store}
-	s.cache.insertRegionToCache(&Region{meta: &r3, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()})
+	s.cache.insertRegionToCache(&Region{meta: &r3, store: unsafe.Pointer(st), lastAccess: time.Now().Unix()}, true)
 	region, err = s.cache.LocateRegionByID(s.bo, s.region)
 	s.Nil(err)
 	s.NotNil(region)