Ranged-export: Remove CachingBlockstore

The improvements in the range-export code lead to avoid reading most blocks
twice, as well as to allowing some blocks to be written to disk multiple times.

The cache hit-rate went down from being close to 50% to a maximum of 12% at
the very end of the export. The reason is that most CIDs are never read twice
since they are correctly tracked in the CID set.

These numbers do not support the maintenance of the CachingBlockstore
code. Additional testing shows that removing it has similar memory-usage
behaviour and about 5 minute-faster execution (around 10%).

Less code to maintain and less options to mess up with.

api/types.go

@@ -403,7 +403,6 @@ func (m *MsgUuidMapType) UnmarshalJSON(b []byte) error {
 type ChainExportConfig struct {
 	WriteBufferSize   int
 	NumWorkers        int
-	CacheSize         int
 	IncludeMessages   bool
 	IncludeReceipts   bool
 	IncludeStateRoots bool

chain/store/snapshot.go

@@ -508,8 +508,7 @@ func (cs *ChainStore) ExportRange(
 	w io.Writer,
 	head, tail *types.TipSet,
 	messages, receipts, stateroots bool,
-	workers int,
-	cacheSize int) error {
+	workers int) error {
 
 	h := &car.CarHeader{
 		Roots:   head.Cids(),
@@ -520,11 +519,6 @@ func (cs *ChainStore) ExportRange(
 		return xerrors.Errorf("failed to write car header: %s", err)
 	}
 
-	cacheStore, err := NewCachingBlockstore(cs.UnionStore(), cacheSize)
-	if err != nil {
-		return err
-	}
-
 	start := time.Now()
 	log.Infow("walking snapshot range",
 		"head", head.Key(),
@@ -544,7 +538,7 @@ func (cs *ChainStore) ExportRange(
 		includeReceipts: receipts,
 	}
 
-	pw, err := newWalkScheduler(ctx, cacheStore, cfg, w)
+	pw, err := newWalkScheduler(ctx, cs.UnionStore(), cfg, w)
 	if err != nil {
 		return err
 	}

cli/chain.go

@@ -1182,11 +1182,6 @@ var ChainExportRangeCmd = &cli.Command{
 			Usage: "specify the number of workers",
 			Value: 1,
 		},
-		&cli.IntFlag{
-			Name:  "cache-size",
-			Usage: "specify the size of the cache (in objects) to use while exporting",
-			Value: 100_000,
-		},
 		&cli.IntFlag{
 			Name:  "write-buffer",
 			Usage: "specify write buffer size",
@@ -1243,7 +1238,6 @@ var ChainExportRangeCmd = &cli.Command{
 			if err := api.ChainExportRangeInternal(ctx, head.Key(), tail.Key(), lapi.ChainExportConfig{
 				WriteBufferSize:   cctx.Int("write-buffer"),
 				NumWorkers:        cctx.Int("workers"),
-				CacheSize:         cctx.Int("cache-size"),
 				IncludeMessages:   cctx.Bool("messages"),
 				IncludeReceipts:   cctx.Bool("receipts"),
 				IncludeStateRoots: cctx.Bool("stateroots"),
@@ -1256,7 +1250,6 @@ var ChainExportRangeCmd = &cli.Command{
 		stream, err := api.ChainExportRange(ctx, head.Key(), tail.Key(), lapi.ChainExportConfig{
 			WriteBufferSize:   cctx.Int("write-buffer"),
 			NumWorkers:        cctx.Int("workers"),
-			CacheSize:         cctx.Int("cache-size"),
 			IncludeMessages:   cctx.Bool("messages"),
 			IncludeReceipts:   cctx.Bool("receipts"),
 			IncludeStateRoots: cctx.Bool("stateroots"),

node/impl/full/chain.go

@@ -630,7 +630,7 @@ func (a ChainAPI) ChainExportRangeInternal(ctx context.Context, head, tail types
 		bw,
 		headTs, tailTs,
 		cfg.IncludeMessages, cfg.IncludeReceipts, cfg.IncludeStateRoots,
-		cfg.NumWorkers, cfg.CacheSize,
+		cfg.NumWorkers,
 	); err != nil {
 		return fmt.Errorf("exporting chain range: %w", err)
 	}
@@ -658,7 +658,7 @@ func (a ChainAPI) ChainExportRange(ctx context.Context, head, tail types.TipSetK
 			headTs,
 			tailTs,
 			cfg.IncludeMessages, cfg.IncludeReceipts, cfg.IncludeStateRoots,
-			cfg.NumWorkers, cfg.CacheSize,
+			cfg.NumWorkers,
 		)
 		bw.Flush()            //nolint:errcheck // it is a write to a pipe
 		w.CloseWithError(err) //nolint:errcheck // it is a pipe