Merge branch 'master' into jen/proof

blockstore/badger/blockstore.go

@@ -8,9 +8,11 @@ import (
 	"path/filepath"
 	"runtime"
 	"sync"
+	"time"
 
 	"github.com/dgraph-io/badger/v2"
 	"github.com/dgraph-io/badger/v2/options"
+	"github.com/dgraph-io/badger/v2/pb"
 	"github.com/multiformats/go-base32"
 	"go.uber.org/zap"
 
@@ -74,20 +76,45 @@ func (b *badgerLogger) Warningf(format string, args ...interface{}) {
 	b.skip2.Warnf(format, args...)
 }
 
+// bsState is the current blockstore state
+type bsState int
+
 const (
-	stateOpen = iota
+	// stateOpen signifies an open blockstore
+	stateOpen bsState = iota
+	// stateClosing signifies a blockstore that is currently closing
 	stateClosing
+	// stateClosed signifies a blockstore that has been colosed
 	stateClosed
 )
 
+type bsMoveState int
+
+const (
+	// moveStateNone signifies that there is no move in progress
+	moveStateNone bsMoveState = iota
+	// moveStateMoving signifies that there is a move  in a progress
+	moveStateMoving
+	// moveStateCleanup signifies that a move has completed or aborted and we are cleaning up
+	moveStateCleanup
+	// moveStateLock signifies that an exclusive lock has been acquired
+	moveStateLock
+)
+
 // Blockstore is a badger-backed IPLD blockstore.
 type Blockstore struct {
 	stateLk sync.RWMutex
-	state   int
+	state   bsState
 	viewers sync.WaitGroup
 
-	DB   *badger.DB
+	moveMx    sync.Mutex
-	opts Options
+	moveCond  sync.Cond
+	moveState bsMoveState
+	rlock     int
+
+	db     *badger.DB
+	dbNext *badger.DB // when moving
+	opts   Options
 
 	prefixing bool
 	prefix    []byte
@@ -113,13 +140,15 @@ func Open(opts Options) (*Blockstore, error) {
 		return nil, fmt.Errorf("failed to open badger blockstore: %w", err)
 	}
 
-	bs := &Blockstore{DB: db, opts: opts}
+	bs := &Blockstore{db: db, opts: opts}
 	if p := opts.Prefix; p != "" {
 		bs.prefixing = true
 		bs.prefix = []byte(p)
 		bs.prefixLen = len(bs.prefix)
 	}
 
+	bs.moveCond.L = &bs.moveMx
+
 	return bs, nil
 }
 
@@ -143,7 +172,7 @@ func (b *Blockstore) Close() error {
 	// wait for all accesses to complete
 	b.viewers.Wait()
 
-	return b.DB.Close()
+	return b.db.Close()
 }
 
 func (b *Blockstore) access() error {
@@ -165,12 +194,225 @@ func (b *Blockstore) isOpen() bool {
 	return b.state == stateOpen
 }
 
-// CollectGarbage runs garbage collection on the value log
+// lockDB/unlockDB implement a recursive lock contingent on move state
-func (b *Blockstore) CollectGarbage() error {
+func (b *Blockstore) lockDB() {
-	if err := b.access(); err != nil {
+	b.moveMx.Lock()
-		return err
+	defer b.moveMx.Unlock()
+
+	if b.rlock == 0 {
+		for b.moveState == moveStateLock {
+			b.moveCond.Wait()
+		}
 	}
-	defer b.viewers.Done()
+
+	b.rlock++
+}
+
+func (b *Blockstore) unlockDB() {
+	b.moveMx.Lock()
+	defer b.moveMx.Unlock()
+
+	b.rlock--
+	if b.rlock == 0 && b.moveState == moveStateLock {
+		b.moveCond.Broadcast()
+	}
+}
+
+// lockMove/unlockMove implement an exclusive lock of move state
+func (b *Blockstore) lockMove() {
+	b.moveMx.Lock()
+	b.moveState = moveStateLock
+	for b.rlock > 0 {
+		b.moveCond.Wait()
+	}
+}
+
+func (b *Blockstore) unlockMove(state bsMoveState) {
+	b.moveState = state
+	b.moveCond.Broadcast()
+	b.moveMx.Unlock()
+}
+
+// movingGC moves the blockstore to a new path, adjacent to the current path, and creates
+// a symlink from the current path to the new path; the old blockstore is deleted.
+//
+// The blockstore MUST accept new writes during the move and ensure that these
+// are persisted to the new blockstore; if a failure occurs aboring the move,
+// then they must be peristed to the old blockstore.
+// In short, the blockstore must not lose data from new writes during the move.
+func (b *Blockstore) movingGC() error {
+	// this inlines moveLock/moveUnlock for the initial state check to prevent a second move
+	// while one is in progress without clobbering state
+	b.moveMx.Lock()
+	if b.moveState != moveStateNone {
+		b.moveMx.Unlock()
+		return fmt.Errorf("move in progress")
+	}
+
+	b.moveState = moveStateLock
+	for b.rlock > 0 {
+		b.moveCond.Wait()
+	}
+
+	b.moveState = moveStateMoving
+	b.moveCond.Broadcast()
+	b.moveMx.Unlock()
+
+	var path string
+
+	defer func() {
+		b.lockMove()
+
+		db2 := b.dbNext
+		b.dbNext = nil
+
+		var state bsMoveState
+		if db2 != nil {
+			state = moveStateCleanup
+		} else {
+			state = moveStateNone
+		}
+
+		b.unlockMove(state)
+
+		if db2 != nil {
+			err := db2.Close()
+			if err != nil {
+				log.Warnf("error closing badger db: %s", err)
+			}
+			b.deleteDB(path)
+
+			b.lockMove()
+			b.unlockMove(moveStateNone)
+		}
+	}()
+
+	// we resolve symlinks to create the new path in the adjacent to the old path.
+	// this allows the user to symlink the db directory into a separate filesystem.
+	basePath := b.opts.Dir
+	linkPath, err := filepath.EvalSymlinks(basePath)
+	if err != nil {
+		return fmt.Errorf("error resolving symlink %s: %w", basePath, err)
+	}
+
+	if basePath == linkPath {
+		path = basePath
+	} else {
+		name := filepath.Base(basePath)
+		dir := filepath.Dir(linkPath)
+		path = filepath.Join(dir, name)
+	}
+	path = fmt.Sprintf("%s.%d", path, time.Now().UnixNano())
+
+	log.Infof("moving blockstore from %s to %s", b.opts.Dir, path)
+
+	opts := b.opts
+	opts.Dir = path
+	opts.ValueDir = path
+
+	db2, err := badger.Open(opts.Options)
+	if err != nil {
+		return fmt.Errorf("failed to open badger blockstore in %s: %w", path, err)
+	}
+
+	b.lockMove()
+	b.dbNext = db2
+	b.unlockMove(moveStateMoving)
+
+	log.Info("copying blockstore")
+	err = b.doCopy(b.db, b.dbNext)
+	if err != nil {
+		return fmt.Errorf("error moving badger blockstore to %s: %w", path, err)
+	}
+
+	b.lockMove()
+	db1 := b.db
+	b.db = b.dbNext
+	b.dbNext = nil
+	b.unlockMove(moveStateCleanup)
+
+	err = db1.Close()
+	if err != nil {
+		log.Warnf("error closing old badger db: %s", err)
+	}
+
+	dbpath := b.opts.Dir
+	oldpath := fmt.Sprintf("%s.old.%d", dbpath, time.Now().Unix())
+
+	if err = os.Rename(dbpath, oldpath); err != nil {
+		// this is not catastrophic in the sense that we have not lost any data.
+		// but it is pretty bad, as the db path points to the old db, while we are now using to the new
+		// db; we can't continue and leave a ticking bomb for the next restart.
+		// so a panic is appropriate and user can fix.
+		panic(fmt.Errorf("error renaming old badger db dir from %s to %s: %w; USER ACTION REQUIRED", dbpath, oldpath, err)) //nolint
+	}
+
+	if err = os.Symlink(path, dbpath); err != nil {
+		// same here; the db path is pointing to the void. panic and let the user fix.
+		panic(fmt.Errorf("error symlinking new badger db dir from %s to %s: %w; USER ACTION REQUIRED", path, dbpath, err)) //nolint
+	}
+
+	b.deleteDB(oldpath)
+
+	log.Info("moving blockstore done")
+	return nil
+}
+
+// doCopy copies a badger blockstore to another, with an optional filter; if the filter
+// is not nil, then only cids that satisfy the filter will be copied.
+func (b *Blockstore) doCopy(from, to *badger.DB) error {
+	workers := runtime.NumCPU() / 2
+	if workers < 2 {
+		workers = 2
+	}
+
+	stream := from.NewStream()
+	stream.NumGo = workers
+	stream.LogPrefix = "doCopy"
+	stream.Send = func(list *pb.KVList) error {
+		batch := to.NewWriteBatch()
+		defer batch.Cancel()
+
+		for _, kv := range list.Kv {
+			if kv.Key == nil || kv.Value == nil {
+				continue
+			}
+			if err := batch.Set(kv.Key, kv.Value); err != nil {
+				return err
+			}
+		}
+
+		return batch.Flush()
+	}
+
+	return stream.Orchestrate(context.Background())
+}
+
+func (b *Blockstore) deleteDB(path string) {
+	// follow symbolic links, otherwise the data wil be left behind
+	lpath, err := filepath.EvalSymlinks(path)
+	if err != nil {
+		log.Warnf("error resolving symlinks in %s", path)
+		return
+	}
+
+	log.Infof("removing data directory %s", lpath)
+	if err := os.RemoveAll(lpath); err != nil {
+		log.Warnf("error deleting db at %s: %s", lpath, err)
+		return
+	}
+
+	if path != lpath {
+		log.Infof("removing link %s", path)
+		if err := os.Remove(path); err != nil {
+			log.Warnf("error removing symbolic link %s", err)
+		}
+	}
+}
+
+func (b *Blockstore) onlineGC() error {
+	b.lockDB()
+	defer b.unlockDB()
 
 	// compact first to gather the necessary statistics for GC
 	nworkers := runtime.NumCPU() / 2
@@ -178,13 +420,13 @@ func (b *Blockstore) CollectGarbage() error {
 		nworkers = 2
 	}
 
-	err := b.DB.Flatten(nworkers)
+	err := b.db.Flatten(nworkers)
 	if err != nil {
 		return err
 	}
 
 	for err == nil {
-		err = b.DB.RunValueLogGC(0.125)
+		err = b.db.RunValueLogGC(0.125)
 	}
 
 	if err == badger.ErrNoRewrite {
@@ -195,6 +437,29 @@ func (b *Blockstore) CollectGarbage() error {
 	return err
 }
 
+// CollectGarbage compacts and runs garbage collection on the value log;
+// implements the BlockstoreGC trait
+func (b *Blockstore) CollectGarbage(opts ...blockstore.BlockstoreGCOption) error {
+	if err := b.access(); err != nil {
+		return err
+	}
+	defer b.viewers.Done()
+
+	var options blockstore.BlockstoreGCOptions
+	for _, opt := range opts {
+		err := opt(&options)
+		if err != nil {
+			return err
+		}
+	}
+
+	if options.FullGC {
+		return b.movingGC()
+	}
+
+	return b.onlineGC()
+}
+
 // Size returns the aggregate size of the blockstore
 func (b *Blockstore) Size() (int64, error) {
 	if err := b.access(); err != nil {
@@ -202,7 +467,10 @@ func (b *Blockstore) Size() (int64, error) {
 	}
 	defer b.viewers.Done()
 
-	lsm, vlog := b.DB.Size()
+	b.lockDB()
+	defer b.unlockDB()
+
+	lsm, vlog := b.db.Size()
 	size := lsm + vlog
 
 	if size == 0 {
@@ -234,12 +502,15 @@ func (b *Blockstore) View(cid cid.Cid, fn func([]byte) error) error {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(cid)
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
-	return b.DB.View(func(txn *badger.Txn) error {
+	return b.db.View(func(txn *badger.Txn) error {
 		switch item, err := txn.Get(k); err {
 		case nil:
 			return item.Value(fn)
@@ -258,12 +529,15 @@ func (b *Blockstore) Has(cid cid.Cid) (bool, error) {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(cid)
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
-	err := b.DB.View(func(txn *badger.Txn) error {
+	err := b.db.View(func(txn *badger.Txn) error {
 		_, err := txn.Get(k)
 		return err
 	})
@@ -289,13 +563,16 @@ func (b *Blockstore) Get(cid cid.Cid) (blocks.Block, error) {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(cid)
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
 	var val []byte
-	err := b.DB.View(func(txn *badger.Txn) error {
+	err := b.db.View(func(txn *badger.Txn) error {
 		switch item, err := txn.Get(k); err {
 		case nil:
 			val, err = item.ValueCopy(nil)
@@ -319,13 +596,16 @@ func (b *Blockstore) GetSize(cid cid.Cid) (int, error) {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(cid)
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
 	var size int
-	err := b.DB.View(func(txn *badger.Txn) error {
+	err := b.db.View(func(txn *badger.Txn) error {
 		switch item, err := txn.Get(k); err {
 		case nil:
 			size = int(item.ValueSize())
@@ -349,18 +629,36 @@ func (b *Blockstore) Put(block blocks.Block) error {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(block.Cid())
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
-	err := b.DB.Update(func(txn *badger.Txn) error {
+	put := func(db *badger.DB) error {
-		return txn.Set(k, block.RawData())
+		err := db.Update(func(txn *badger.Txn) error {
-	})
+			return txn.Set(k, block.RawData())
-	if err != nil {
+		})
-		err = fmt.Errorf("failed to put block in badger blockstore: %w", err)
+		if err != nil {
+			return fmt.Errorf("failed to put block in badger blockstore: %w", err)
+		}
+
+		return nil
 	}
-	return err
+
+	if err := put(b.db); err != nil {
+		return err
+	}
+
+	if b.dbNext != nil {
+		if err := put(b.dbNext); err != nil {
+			return err
+		}
+	}
+
+	return nil
 }
 
 // PutMany implements Blockstore.PutMany.
@@ -370,6 +668,9 @@ func (b *Blockstore) PutMany(blocks []blocks.Block) error {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	// toReturn tracks the byte slices to return to the pool, if we're using key
 	// prefixing. we can't return each slice to the pool after each Set, because
 	// badger holds on to the slice.
@@ -383,24 +684,45 @@ func (b *Blockstore) PutMany(blocks []blocks.Block) error {
 		}()
 	}
 
-	batch := b.DB.NewWriteBatch()
+	keys := make([][]byte, 0, len(blocks))
-	defer batch.Cancel()
-
 	for _, block := range blocks {
 		k, pooled := b.PooledStorageKey(block.Cid())
 		if pooled {
 			toReturn = append(toReturn, k)
 		}
-		if err := batch.Set(k, block.RawData()); err != nil {
+		keys = append(keys, k)
+	}
+
+	put := func(db *badger.DB) error {
+		batch := db.NewWriteBatch()
+		defer batch.Cancel()
+
+		for i, block := range blocks {
+			k := keys[i]
+			if err := batch.Set(k, block.RawData()); err != nil {
+				return err
+			}
+		}
+
+		err := batch.Flush()
+		if err != nil {
+			return fmt.Errorf("failed to put blocks in badger blockstore: %w", err)
+		}
+
+		return nil
+	}
+
+	if err := put(b.db); err != nil {
+		return err
+	}
+
+	if b.dbNext != nil {
+		if err := put(b.dbNext); err != nil {
 			return err
 		}
 	}
 
-	err := batch.Flush()
+	return nil
-	if err != nil {
-		err = fmt.Errorf("failed to put blocks in badger blockstore: %w", err)
-	}
-	return err
 }
 
 // DeleteBlock implements Blockstore.DeleteBlock.
@@ -410,12 +732,15 @@ func (b *Blockstore) DeleteBlock(cid cid.Cid) error {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	k, pooled := b.PooledStorageKey(cid)
 	if pooled {
 		defer KeyPool.Put(k)
 	}
 
-	return b.DB.Update(func(txn *badger.Txn) error {
+	return b.db.Update(func(txn *badger.Txn) error {
 		return txn.Delete(k)
 	})
 }
@@ -426,6 +751,9 @@ func (b *Blockstore) DeleteMany(cids []cid.Cid) error {
 	}
 	defer b.viewers.Done()
 
+	b.lockDB()
+	defer b.unlockDB()
+
 	// toReturn tracks the byte slices to return to the pool, if we're using key
 	// prefixing. we can't return each slice to the pool after each Set, because
 	// badger holds on to the slice.
@@ -439,7 +767,7 @@ func (b *Blockstore) DeleteMany(cids []cid.Cid) error {
 		}()
 	}
 
-	batch := b.DB.NewWriteBatch()
+	batch := b.db.NewWriteBatch()
 	defer batch.Cancel()
 
 	for _, cid := range cids {
@@ -465,7 +793,10 @@ func (b *Blockstore) AllKeysChan(ctx context.Context) (<-chan cid.Cid, error) {
 		return nil, err
 	}
 
-	txn := b.DB.NewTransaction(false)
+	b.lockDB()
+	defer b.unlockDB()
+
+	txn := b.db.NewTransaction(false)
 	opts := badger.IteratorOptions{PrefetchSize: 100}
 	if b.prefixing {
 		opts.Prefix = b.prefix
@@ -519,7 +850,10 @@ func (b *Blockstore) ForEachKey(f func(cid.Cid) error) error {
 	}
 	defer b.viewers.Done()
 
-	txn := b.DB.NewTransaction(false)
+	b.lockDB()
+	defer b.unlockDB()
+
+	txn := b.db.NewTransaction(false)
 	defer txn.Discard()
 
 	opts := badger.IteratorOptions{PrefetchSize: 100}
@@ -614,3 +948,9 @@ func (b *Blockstore) StorageKey(dst []byte, cid cid.Cid) []byte {
 	}
 	return dst[:reqsize]
 }
+
+// this method is added for lotus-shed needs
+// WARNING: THIS IS COMPLETELY UNSAFE; DONT USE THIS IN PRODUCTION CODE
+func (b *Blockstore) DB() *badger.DB {
+	return b.db
+}

blockstore/badger/blockstore_test.go

@@ -1,12 +1,19 @@
 package badgerbs
 
 import (
+	"bytes"
+	"fmt"
 	"io/ioutil"
 	"os"
+	"path/filepath"
+	"strings"
 	"testing"
 
-	blocks "github.com/ipfs/go-block-format"
 	"github.com/stretchr/testify/require"
+	"golang.org/x/sync/errgroup"
+
+	blocks "github.com/ipfs/go-block-format"
+	cid "github.com/ipfs/go-cid"
 
 	"github.com/filecoin-project/lotus/blockstore"
 )
@@ -89,3 +96,165 @@ func openBlockstore(optsSupplier func(path string) Options) func(tb testing.TB,
 		return Open(optsSupplier(path))
 	}
 }
+
+func testMove(t *testing.T, optsF func(string) Options) {
+	basePath, err := ioutil.TempDir("", "")
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	dbPath := filepath.Join(basePath, "db")
+
+	t.Cleanup(func() {
+		_ = os.RemoveAll(basePath)
+	})
+
+	db, err := Open(optsF(dbPath))
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	defer db.Close() //nolint
+
+	var have []blocks.Block
+	var deleted []cid.Cid
+
+	// add some blocks
+	for i := 0; i < 10; i++ {
+		blk := blocks.NewBlock([]byte(fmt.Sprintf("some data %d", i)))
+		err := db.Put(blk)
+		if err != nil {
+			t.Fatal(err)
+		}
+		have = append(have, blk)
+	}
+
+	// delete some of them
+	for i := 5; i < 10; i++ {
+		c := have[i].Cid()
+		err := db.DeleteBlock(c)
+		if err != nil {
+			t.Fatal(err)
+		}
+		deleted = append(deleted, c)
+	}
+	have = have[:5]
+
+	// start a move concurrent with some more puts
+	g := new(errgroup.Group)
+	g.Go(func() error {
+		for i := 10; i < 1000; i++ {
+			blk := blocks.NewBlock([]byte(fmt.Sprintf("some data %d", i)))
+			err := db.Put(blk)
+			if err != nil {
+				return err
+			}
+			have = append(have, blk)
+		}
+		return nil
+	})
+	g.Go(func() error {
+		return db.CollectGarbage(blockstore.WithFullGC(true))
+	})
+
+	err = g.Wait()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	// now check that we have all the blocks in have and none in the deleted lists
+	checkBlocks := func() {
+		for _, blk := range have {
+			has, err := db.Has(blk.Cid())
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			if !has {
+				t.Fatal("missing block")
+			}
+
+			blk2, err := db.Get(blk.Cid())
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			if !bytes.Equal(blk.RawData(), blk2.RawData()) {
+				t.Fatal("data mismatch")
+			}
+		}
+
+		for _, c := range deleted {
+			has, err := db.Has(c)
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			if has {
+				t.Fatal("resurrected block")
+			}
+		}
+	}
+
+	checkBlocks()
+
+	// check the basePath -- it should contain a directory with name db.{timestamp}, soft-linked
+	// to db and nothing else
+	checkPath := func() {
+		entries, err := os.ReadDir(basePath)
+		if err != nil {
+			t.Fatal(err)
+		}
+
+		if len(entries) != 2 {
+			t.Fatalf("too many entries; expected %d but got %d", 2, len(entries))
+		}
+
+		var haveDB, haveDBLink bool
+		for _, e := range entries {
+			if e.Name() == "db" {
+				if (e.Type() & os.ModeSymlink) == 0 {
+					t.Fatal("found db, but it's not a symlink")
+				}
+				haveDBLink = true
+				continue
+			}
+			if strings.HasPrefix(e.Name(), "db.") {
+				if !e.Type().IsDir() {
+					t.Fatal("found db prefix, but it's not a directory")
+				}
+				haveDB = true
+				continue
+			}
+		}
+
+		if !haveDB {
+			t.Fatal("db directory is missing")
+		}
+		if !haveDBLink {
+			t.Fatal("db link is missing")
+		}
+	}
+
+	checkPath()
+
+	// now do another FullGC to test the double move and following of symlinks
+	if err := db.CollectGarbage(blockstore.WithFullGC(true)); err != nil {
+		t.Fatal(err)
+	}
+
+	checkBlocks()
+	checkPath()
+}
+
+func TestMoveNoPrefix(t *testing.T) {
+	testMove(t, DefaultOptions)
+}
+
+func TestMoveWithPrefix(t *testing.T) {
+	testMove(t, func(path string) Options {
+		opts := DefaultOptions(path)
+		opts.Prefix = "/prefixed/"
+		return opts
+	})
+}

blockstore/blockstore.go

@@ -37,7 +37,22 @@ type BlockstoreIterator interface {
 
 // BlockstoreGC is a trait for blockstores that support online garbage collection
 type BlockstoreGC interface {
-	CollectGarbage() error
+	CollectGarbage(options ...BlockstoreGCOption) error
+}
+
+// BlockstoreGCOption is a functional interface for controlling blockstore GC options
+type BlockstoreGCOption = func(*BlockstoreGCOptions) error
+
+// BlockstoreGCOptions is a struct with GC options
+type BlockstoreGCOptions struct {
+	FullGC bool
+}
+
+func WithFullGC(fullgc bool) BlockstoreGCOption {
+	return func(opts *BlockstoreGCOptions) error {
+		opts.FullGC = fullgc
+		return nil
+	}
 }
 
 // BlockstoreSize is a trait for on-disk blockstores that can report their size

blockstore/splitstore/README.md

@@ -59,6 +59,15 @@ These are options in the `[Chainstore.Splitstore]` section of the configuration:
   nodes beyond 4 finalities, while running with the discard coldstore option.
   It is also useful for miners who accept deals and need to lookback messages beyond
   the 4 finalities, which would otherwise hit the coldstore.
+- `HotStoreFullGCFrequency` -- specifies how frequenty to garbage collect the hotstore
+  using full (moving) GC.
+  The default value is 20, which uses full GC every 20 compactions (about once a week);
+  set to 0 to disable full GC altogether.
+  Rationale: badger supports online GC, and this is used by default. However it has proven to
+  be ineffective in practice with the hotstore size slowly creeping up. In order to address this,
+  we have added moving GC support in our badger wrapper, which can effectively reclaim all space.
+  The downside is that it takes a bit longer to perform a moving GC and you also need enough
+  space to house the new hotstore while the old one is still live.
 
 
 ## Operation

blockstore/splitstore/splitstore.go

@@ -81,6 +81,13 @@ type Config struct {
 	// - a positive integer indicates the number of finalities, outside the compaction boundary,
 	//   for which messages will be retained in the hotstore.
 	HotStoreMessageRetention uint64
+
+	// HotstoreFullGCFrequency indicates how frequently (in terms of compactions) to garbage collect
+	// the hotstore using full (moving) GC if supported by the hotstore.
+	// A value of 0 disables full GC entirely.
+	// A positive value is the number of compactions before a full GC is performed;
+	// a value of 1 will perform full GC in every compaction.
+	HotStoreFullGCFrequency uint64
 }
 
 // ChainAccessor allows the Splitstore to access the chain. It will most likely

blockstore/splitstore/splitstore_gc.go

@@ -8,17 +8,22 @@ import (
 )
 
 func (s *SplitStore) gcHotstore() {
-	if err := s.gcBlockstoreOnline(s.hot); err != nil {
+	var opts []bstore.BlockstoreGCOption
+	if s.cfg.HotStoreFullGCFrequency > 0 && s.compactionIndex%int64(s.cfg.HotStoreFullGCFrequency) == 0 {
+		opts = append(opts, bstore.WithFullGC(true))
+	}
+
+	if err := s.gcBlockstore(s.hot, opts); err != nil {
 		log.Warnf("error garbage collecting hostore: %s", err)
 	}
 }
 
-func (s *SplitStore) gcBlockstoreOnline(b bstore.Blockstore) error {
+func (s *SplitStore) gcBlockstore(b bstore.Blockstore, opts []bstore.BlockstoreGCOption) error {
 	if gc, ok := b.(bstore.BlockstoreGC); ok {
 		log.Info("garbage collecting blockstore")
 		startGC := time.Now()
 
-		if err := gc.CollectGarbage(); err != nil {
+		if err := gc.CollectGarbage(opts...); err != nil {
 			return err
 		}
 
@@ -26,5 +31,5 @@ func (s *SplitStore) gcBlockstoreOnline(b bstore.Blockstore) error {
 		return nil
 	}
 
-	return fmt.Errorf("blockstore doesn't support online gc: %T", b)
+	return fmt.Errorf("blockstore doesn't support garbage collection: %T", b)
 }

cmd/lotus-shed/pruning.go

@@ -161,7 +161,7 @@ var stateTreePruneCmd = &cli.Command{
 		if cctx.Bool("only-ds-gc") {
 			fmt.Println("running datastore gc....")
 			for i := 0; i < cctx.Int("gc-count"); i++ {
-				if err := badgbs.DB.RunValueLogGC(DiscardRatio); err != nil {
+				if err := badgbs.DB().RunValueLogGC(DiscardRatio); err != nil {
 					return xerrors.Errorf("datastore GC failed: %w", err)
 				}
 			}
@@ -208,7 +208,7 @@ var stateTreePruneCmd = &cli.Command{
 			return nil
 		}
 
-		b := badgbs.DB.NewWriteBatch()
+		b := badgbs.DB().NewWriteBatch()
 		defer b.Cancel()
 
 		markForRemoval := func(c cid.Cid) error {
@@ -249,7 +249,7 @@ var stateTreePruneCmd = &cli.Command{
 
 		fmt.Println("running datastore gc....")
 		for i := 0; i < cctx.Int("gc-count"); i++ {
-			if err := badgbs.DB.RunValueLogGC(DiscardRatio); err != nil {
+			if err := badgbs.DB().RunValueLogGC(DiscardRatio); err != nil {
 				return xerrors.Errorf("datastore GC failed: %w", err)
 			}
 		}

node/config/def.go

@@ -72,6 +72,8 @@ func DefaultFullNode() *FullNode {
 				ColdStoreType: "universal",
 				HotStoreType:  "badger",
 				MarkSetType:   "map",
+
+				HotStoreFullGCFrequency: 20,
 			},
 		},
 	}

node/config/doc_gen.go

@@ -685,25 +685,37 @@ submitting proofs to the chain individually`,
 			Name: "ColdStoreType",
 			Type: "string",
 
-			Comment: ``,
+			Comment: `ColdStoreType specifies the type of the coldstore.
+It can be "universal" (default) or "discard" for discarding cold blocks.`,
 		},
 		{
 			Name: "HotStoreType",
 			Type: "string",
 
-			Comment: ``,
+			Comment: `HotStoreType specifies the type of the hotstore.
+Only currently supported value is "badger".`,
 		},
 		{
 			Name: "MarkSetType",
 			Type: "string",
 
-			Comment: ``,
+			Comment: `MarkSetType specifies the type of the markset.
+It can be "map" (default) for in memory marking or "badger" for on-disk marking.`,
 		},
 		{
 			Name: "HotStoreMessageRetention",
 			Type: "uint64",
 
-			Comment: ``,
+			Comment: `HotStoreMessageRetention specifies the retention policy for messages, in finalities beyond
+the compaction boundary; default is 0.`,
+		},
+		{
+			Name: "HotStoreFullGCFrequency",
+			Type: "uint64",
+
+			Comment: `HotStoreFullGCFrequency specifies how often to perform a full (moving) GC on the hotstore.
+A value of 0 disables, while a value 1 will do full GC in every compaction.
+Default is 20 (about once a week).`,
 		},
 	},
 	"StorageMiner": []DocField{

node/config/types.go

@@ -278,11 +278,23 @@ type Chainstore struct {
 }
 
 type Splitstore struct {
+	// ColdStoreType specifies the type of the coldstore.
+	// It can be "universal" (default) or "discard" for discarding cold blocks.
 	ColdStoreType string
-	HotStoreType  string
+	// HotStoreType specifies the type of the hotstore.
-	MarkSetType   string
+	// Only currently supported value is "badger".
+	HotStoreType string
+	// MarkSetType specifies the type of the markset.
+	// It can be "map" (default) for in memory marking or "badger" for on-disk marking.
+	MarkSetType string
 
+	// HotStoreMessageRetention specifies the retention policy for messages, in finalities beyond
+	// the compaction boundary; default is 0.
 	HotStoreMessageRetention uint64
+	// HotStoreFullGCFrequency specifies how often to perform a full (moving) GC on the hotstore.
+	// A value of 0 disables, while a value 1 will do full GC in every compaction.
+	// Default is 20 (about once a week).
+	HotStoreFullGCFrequency uint64
 }
 
 // // Full Node

node/modules/blockstore.go

@@ -81,6 +81,7 @@ func SplitBlockstore(cfg *config.Chainstore) func(lc fx.Lifecycle, r repo.Locked
 			MarkSetType:              cfg.Splitstore.MarkSetType,
 			DiscardColdBlocks:        cfg.Splitstore.ColdStoreType == "discard",
 			HotStoreMessageRetention: cfg.Splitstore.HotStoreMessageRetention,
+			HotStoreFullGCFrequency:  cfg.Splitstore.HotStoreFullGCFrequency,
 		}
 		ss, err := splitstore.Open(path, ds, hot, cold, cfg)
 		if err != nil {