feat: switch default database backend from goleveldb to pebble

cmd/cometbft/commands/compact.go

@@ -1,41 +1,53 @@
 package commands
 
 import (
-	"errors"
+	"fmt"
+	"os"
 	"path/filepath"
 	"sync"
 
+	"github.com/cockroachdb/pebble"
 	"github.com/spf13/cobra"
 	"github.com/syndtr/goleveldb/leveldb"
 	"github.com/syndtr/goleveldb/leveldb/opt"
 	"github.com/syndtr/goleveldb/leveldb/util"
 
+	cfg "github.com/tendermint/tendermint/config"
 	"github.com/tendermint/tendermint/libs/log"
 )
 
-var CompactGoLevelDBCmd = &cobra.Command{
-	Use:     "experimental-compact-goleveldb",
-	Aliases: []string{"experimental_compact_goleveldb"},
-	Short:   "force compacts the CometBFT storage engine (only GoLevelDB supported)",
+var CompactCmd = &cobra.Command{
+	Use:     "experimental-compact",
+	Aliases: []string{"experimental_compact"},
+	Short:   "force compacts the CometBFT storage engine (supports GoLevelDB and Pebble)",
 	Long: `
-This is a temporary utility command that performs a force compaction on the state
-and blockstores to reduce disk space for a pruning node. This should only be run
-once the node has stopped. This command will likely be omitted in the future after
-the planned refactor to the storage engine.
+This command forces a compaction of the CometBFT storage engine.
 
-Currently, only GoLevelDB is supported.
-	`,
+Currently, GoLevelDB and Pebble are supported.
+`,
 	RunE: func(cmd *cobra.Command, args []string) error {
-		if config.DBBackend != "goleveldb" {
-			return errors.New("compaction is currently only supported with goleveldb")
+		configFile, err := cmd.Flags().GetString("config")
+		if err != nil {
+			return err
 		}
 
-		compactGoLevelDBs(config.RootDir, logger)
-		return nil
+		conf := cfg.DefaultConfig()
+		conf.SetRoot(configFile)
+
+		logger := log.NewTMLogger(log.NewSyncWriter(os.Stdout))
+
+		switch conf.DBBackend {
+		case "goleveldb":
+			return compactGoLevelDBs(conf.RootDir, logger)
+		case "pebble":
+			return compactPebbleDBs(conf.RootDir, logger)
+		default:
+			return fmt.Errorf("compaction is currently only supported with goleveldb and pebble, got %s", conf.DBBackend)
+		}
 	},
 }
 
-func compactGoLevelDBs(rootDir string, logger log.Logger) {
+func compactGoLevelDBs(rootDir string, logger log.Logger) error {
 	dbNames := []string{"state", "blockstore"}
 	o := &opt.Options{
 		DisableSeeksCompaction: true,
@@ -64,4 +76,30 @@ func compactGoLevelDBs(rootDir string, logger log.Logger) {
 		}()
 	}
 	wg.Wait()
+	return nil
+}
+
+func compactPebbleDBs(rootDir string, logger log.Logger) error {
+	dbNames := []string{"state", "blockstore", "evidence", "tx_index"}
+	for _, dbName := range dbNames {
+		dbPath := filepath.Join(rootDir, "data", dbName+".db")
+		logger.Info("Compacting db", "path", dbPath)
+
+		db, err := pebble.Open(dbPath, &pebble.Options{})
+		if err != nil {
+			return fmt.Errorf("failed to open db %s: %w", dbPath, err)
+		}
+
+		err = db.Compact(nil, nil, true)
+		if err != nil {
+			db.Close()
+			return fmt.Errorf("failed to compact db %s: %w", dbPath, err)
+		}
+
+		err = db.Close()
+		if err != nil {
+			return fmt.Errorf("failed to close db %s: %w", dbPath, err)
+		}
+	}
+	return nil
 }

cmd/cometbft/main.go

@@ -30,7 +30,7 @@ func main() {
 		cmd.GenNodeKeyCmd,
 		cmd.VersionCmd,
 		cmd.RollbackStateCmd,
-		cmd.CompactGoLevelDBCmd,
+		cmd.CompactCmd,
 		debug.DebugCmd,
 		cli.NewCompletionCmd(rootCmd, true),
 	)

config/config.go

@@ -182,25 +182,13 @@ type BaseConfig struct { //nolint: maligned
 	// and verifying their commits
 	FastSyncMode bool `mapstructure:"fast_sync"`
 
-	// Database backend: goleveldb | cleveldb | boltdb | rocksdb
+	// Database backend: goleveldb | cleveldb | boltdb | rocksdb | pebble
 	// * goleveldb (github.com/syndtr/goleveldb - most popular implementation)
-	//   - pure go
-	//   - stable
 	// * cleveldb (uses levigo wrapper)
-	//   - fast
-	//   - requires gcc
-	//   - use cleveldb build tag (go build -tags cleveldb)
 	// * boltdb (uses etcd's fork of bolt - github.com/etcd-io/bbolt)
-	//   - EXPERIMENTAL
-	//   - may be faster is some use-cases (random reads - indexer)
-	//   - use boltdb build tag (go build -tags boltdb)
-	// * rocksdb (uses github.com/tecbot/gorocksdb)
-	//   - EXPERIMENTAL
-	//   - requires gcc
-	//   - use rocksdb build tag (go build -tags rocksdb)
+	// * rocksdb (uses github.com/linxGnu/grocksdb)
+	// * pebble (uses github.com/cockroachdb/pebble - most performant implementation)
 	// * badgerdb (uses github.com/dgraph-io/badger)
-	//   - EXPERIMENTAL
-	//   - use badgerdb build tag (go build -tags badgerdb)
 	DBBackend string `mapstructure:"db_backend"`
 
 	// Database directory
@@ -250,7 +238,7 @@ func DefaultBaseConfig() BaseConfig {
 		LogFormat:          LogFormatPlain,
 		FastSyncMode:       true,
 		FilterPeers:        false,
-		DBBackend:          "goleveldb",
+		DBBackend:          "pebble",
 		DBPath:             "data",
 	}
 }
@@ -531,7 +519,7 @@ type P2PConfig struct { //nolint: maligned
 	ExternalAddress string `mapstructure:"external_address"`
 
 	// Comma separated list of seed nodes to connect to
-	// We only use these if we can’t connect to peers in the addrbook
+	// We only use these if we can't connect to peers in the addrbook
 	Seeds string `mapstructure:"seeds"`
 
 	// Comma separated list of nodes to keep persistent connections to

config/toml.go

@@ -92,26 +92,14 @@ moniker = "{{ .BaseConfig.Moniker }}"
 # and verifying their commits
 fast_sync = {{ .BaseConfig.FastSyncMode }}
 
-# Database backend: goleveldb | cleveldb | boltdb | rocksdb | badgerdb
+# Database backend: goleveldb | cleveldb | boltdb | rocksdb | badgerdb | pebble
 # * goleveldb (github.com/syndtr/goleveldb - most popular implementation)
-#   - pure go
-#   - stable
 # * cleveldb (uses levigo wrapper)
-#   - fast
-#   - requires gcc
-#   - use cleveldb build tag (go build -tags cleveldb)
 # * boltdb (uses etcd's fork of bolt - github.com/etcd-io/bbolt)
-#   - EXPERIMENTAL
-#   - may be faster is some use-cases (random reads - indexer)
-#   - use boltdb build tag (go build -tags boltdb)
-# * rocksdb (uses github.com/tecbot/gorocksdb)
-#   - EXPERIMENTAL
-#   - requires gcc
-#   - use rocksdb build tag (go build -tags rocksdb)
+# * rocksdb (uses github.com/linxGnu/grocksdb)
+# * pebble (uses github.com/cockroachdb/pebble - most performant implementation)
 # * badgerdb (uses github.com/dgraph-io/badger)
-#   - EXPERIMENTAL
-#   - use badgerdb build tag (go build -tags badgerdb)
-db_backend = "{{ .BaseConfig.DBBackend }}"
+db_backend = "pebble"
 
 # Database directory
 db_dir = "{{ js .BaseConfig.DBPath }}"

test/e2e/generator/generate.go

@@ -33,7 +33,7 @@ var (
 	}
 
 	// The following specify randomly chosen values for testnet nodes.
-	nodeDatabases = uniformChoice{"goleveldb"}
+	nodeDatabases = uniformChoice{"pebble"}
 	ipv6          = uniformChoice{false, true}
 	// FIXME: grpc disabled due to https://github.com/tendermint/tendermint/issues/5439
 	nodeABCIProtocols     = uniformChoice{"unix", "tcp", "builtin"}

test/e2e/pkg/testnet.go

@@ -222,7 +222,7 @@ func NewTestnetFromManifest(manifest Manifest, file string, ifd InfrastructureDa
 			ExternalIP:       extIP,
 			ProxyPort:        ind.Port,
 			Mode:             ModeValidator,
-			Database:         "goleveldb",
+			Database:         "pebble",
 			ABCIProtocol:     Protocol(testnet.ABCIProtocol),
 			PrivvalProtocol:  ProtocolFile,
 			StartAt:          nodeManifest.StartAt,
@@ -393,7 +393,7 @@ func (n Node) Validate(testnet Testnet) error {
 		return fmt.Errorf("invalid block sync setting %q", n.BlockSyncVersion)
 	}
 	switch n.Database {
-	case "goleveldb":
+	case "pebble":
 	default:
 		return fmt.Errorf("invalid database setting %q", n.Database)
 	}

test/loadtime/README.md

@@ -47,11 +47,11 @@ the timestamp of the block the transaction was executed in to determine transact
 minimum, maximum, and average latency as well as the standard deviation.
 
 Below is a basic invocation of the report tool with a data directory under `/home/test/.cometbft/data/`
-where the data was saved in a `goleveldb` database.
+where the data was saved in a `pebble` database.
 
 
 ```bash
-./build/report --database-type goleveldb --data-dir ~/.cometbft/data
+./build/report --database-type pebble --data-dir ~/.cometbft/data
 ```
 
 The `report` tool also supports outputting the raw data as `csv`. This can be
@@ -61,7 +61,7 @@ Below is an invocation of the report tool that outputs the data to a `csv` file
 in `out.csv`
 
 ```bash
-./build/report --database-type goleveldb --data-dir ~/.cometbft/data --csv out.csv
+./build/report --database-type pebble --data-dir ~/.cometbft/data --csv out.csv
 ```
 
 The `report` tool outputs the data for each experiment separately, identified

test/loadtime/cmd/report/main.go

@@ -16,7 +16,7 @@ import (
 )
 
 var (
-	db     = flag.String("database-type", "goleveldb", "the type of database holding the blockstore")
+	db     = flag.String("database-type", "pebble", "the type of database holding the blockstore")
 	dir    = flag.String("data-dir", "", "path to the directory containing the CometBFT databases")
 	csvOut = flag.String("csv", "", "dump the extracted latencies as raw csv for use in additional tooling")
 )