Frrist/meta/instanceid (#4411)

Co-authored-by: frrist <forrest@expanso.io>

go.mod

@@ -13,6 +13,7 @@ require (
 	github.com/bmatcuk/doublestar/v4 v4.6.1
 	github.com/c2h5oh/datasize v0.0.0-20220606134207-859f65c6625b
 	github.com/davecgh/go-spew v1.1.2-0.20180830191138-d8f796af33cc
+	github.com/denisbrodbeck/machineid v1.0.1
 	github.com/docker/docker v27.1.1+incompatible
 	github.com/dylibso/observe-sdk/go v0.0.0-20231201014635-141351c24659
 	github.com/fatih/structs v1.1.0

go.sum

@@ -401,6 +401,8 @@ github.com/decred/dcrd/crypto/blake256 v1.0.1 h1:7PltbUIQB7u/FfZ39+DGa/ShuMyJ5il
 github.com/decred/dcrd/crypto/blake256 v1.0.1/go.mod h1:2OfgNZ5wDpcsFmHmCK5gZTPcCXqlm2ArzUIkw9czNJo=
 github.com/decred/dcrd/dcrec/secp256k1/v4 v4.2.0 h1:8UrgZ3GkP4i/CLijOJx79Yu+etlyjdBU4sfcs2WYQMs=
 github.com/decred/dcrd/dcrec/secp256k1/v4 v4.2.0/go.mod h1:v57UDF4pDQJcEfFUCRop3lJL149eHGSe9Jvczhzjo/0=
+github.com/denisbrodbeck/machineid v1.0.1 h1:geKr9qtkB876mXguW2X6TU4ZynleN6ezuMSRhl4D7AQ=
+github.com/denisbrodbeck/machineid v1.0.1/go.mod h1:dJUwb7PTidGDeYyUBmXZ2GphQBbjJCrnectwCyxcUSI=
 github.com/dgraph-io/badger v1.6.2 h1:mNw0qs90GVgGGWylh0umH5iag1j6n/PeJtNvL6KY/x8=
 github.com/dgraph-io/badger v1.6.2/go.mod h1:JW2yswe3V058sS0kZ2h/AXeDSqFjxnZcRrVH//y2UQE=
 github.com/dgraph-io/badger/v3 v3.2103.5 h1:ylPa6qzbjYRQMU6jokoj4wzcaweHylt//CH0AKt0akg=

pkg/repo/fs.go

@@ -98,6 +98,11 @@ func (fsr *FsRepo) Init() error {
 	// TODO this should be a part of the config.
 	telemetry.SetupFromEnvs()
 
+	// never fail here as this isn't critical to node start up.
+	if err := fsr.WriteInstanceID(GenerateInstanceID()); err != nil {
+		log.Trace().Err(err).Msgf("failed to write instanceID")
+	}
+
 	if err := fsr.WriteVersion(Version4); err != nil {
 		return fmt.Errorf("failed to persist repo version: %w", err)
 	}
@@ -120,6 +125,16 @@ func (fsr *FsRepo) Open() error {
 		}
 	}
 
+	// check if an instanceID exists persisting one if not found.
+	// never fail here as this isn't critical to node start up.
+	if instanceID, err := fsr.ReadInstanceID(); err != nil {
+		log.Trace().Err(err).Msgf("failed to read instanceID")
+	} else if instanceID == "" {
+		if err := fsr.WriteInstanceID(GenerateInstanceID()); err != nil {
+			log.Trace().Err(err).Msgf("failed to write instanceID")
+		}
+	}
+
 	// TODO this should be a part of the config.
 	telemetry.SetupFromEnvs()
 

pkg/repo/instanceid.go

@@ -0,0 +1,24 @@
+package repo
+
+import (
+	"crypto/sha256"
+	"encoding/hex"
+
+	"github.com/denisbrodbeck/machineid"
+	"github.com/google/uuid"
+)
+
+// GenerateInstanceID creates a unique, anonymous identifier for the instance of bacalhau.
+func GenerateInstanceID() string {
+	// Get machine ID, which provides a consistent identifier for the device
+	machineID, err := machineid.ID()
+	if err != nil {
+		// if we fail to read a machineID, generate a UUID
+		machineID = uuid.NewString()
+	}
+
+	// hash the machineID to make it anonymous.
+	hash := sha256.New()
+	hash.Write([]byte(machineID))
+	return hex.EncodeToString(hash.Sum(nil))
+}

pkg/repo/sysmeta.go

@@ -34,6 +34,7 @@ const SystemMetadataFile = "system_metadata.yaml"
 type SystemMetadata struct {
 	RepoVersion     int       `yaml:"RepoVersion"`
 	InstallationID  string    `yaml:"InstallationID"`
+	InstanceID      string    `yaml:"InstanceID"`
 	LastUpdateCheck time.Time `yaml:"LastUpdateCheck"`
 	NodeName        string    `yaml:"NodeName"`
 }
@@ -98,6 +99,24 @@ func (fsr *FsRepo) WriteInstallationID(id string) error {
 	})
 }
 
+// ReadInstanceID reads the InstanceID in the metadata.
+// It fails if the metadata file doesn't exist.
+func (fsr *FsRepo) ReadInstanceID() (string, error) {
+	sysmeta, err := fsr.readMetadata()
+	if err != nil {
+		return "", err
+	}
+	return sysmeta.InstanceID, nil
+}
+
+// WriteInstanceID updates the InstanceID in the metadata.
+// It fails if the metadata file doesn't exist.
+func (fsr *FsRepo) WriteInstanceID(id string) error {
+	return fsr.updateExistingMetadata(func(sysmeta *SystemMetadata) {
+		sysmeta.InstanceID = id
+	})
+}
+
 func (fsr *FsRepo) ReadNodeName() (string, error) {
 	sysmeta, err := fsr.readMetadata()
 	if err != nil {