feat(issue-253): initial metrics implementation

README.md

@@ -345,6 +345,10 @@ sequenceDiagram
     mev_boost-->>consensus: submitBlindedBlock response
 ```
 
+# Metrics
+
+Optionally the `-prometheus-port=9000` flag can be passed to expose prometheus metrics on specified port
+
 # Maintainers
 
 - [@metachris](https://github.com/metachris)

cli/flags.go

@@ -34,6 +34,7 @@ var flags = []cli.Flag{
 	timeoutGetPayloadFlag,
 	timeoutRegValFlag,
 	maxRetriesFlag,
+	prometheusListenAddr,
 }
 
 var (
@@ -83,6 +84,12 @@ var (
 		Usage:    "disables adding the version to every log entry",
 		Category: LoggingCategory,
 	}
+	prometheusListenAddr = &cli.IntFlag{
+		Name:     "prometheus-port",
+		Sources:  cli.EnvVars("PROMETHEUS_PORT"),
+		Usage:    "when set to a valid http port, will export runtime metrics to a prometheus server on that port",
+		Category: LoggingCategory,
+	}
 	// Genesis Flags
 	customGenesisForkFlag = &cli.StringFlag{
 		Name:     "genesis-fork-version",

cli/main.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"flag"
 	"fmt"
+	"math"
 	"os"
 	"strings"
 	"time"
@@ -13,6 +14,8 @@ import (
 	"github.com/flashbots/mev-boost/common"
 	"github.com/flashbots/mev-boost/config"
 	"github.com/flashbots/mev-boost/server"
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/collectors"
 	"github.com/sirupsen/logrus"
 	"github.com/urfave/cli/v3"
 )
@@ -69,6 +72,13 @@ func start(_ context.Context, cmd *cli.Command) error {
 		relays, monitors, minBid, relayCheck = setupRelays(cmd)
 		listenAddr                           = cmd.String(addrFlag.Name)
 	)
+	prometheusRegistry := prometheus.NewRegistry()
+	if err := prometheusRegistry.Register(collectors.NewGoCollector()); err != nil {
+		log.WithError(err).Error("Failed to register metrics for GoCollector")
+	}
+	if err := prometheusRegistry.Register(collectors.NewProcessCollector(collectors.ProcessCollectorOpts{})); err != nil {
+		log.WithError(err).Error("Failed to register ProcessCollector")
+	}
 
 	opts := server.BoostServiceOpts{
 		Log:                      log,
@@ -83,6 +93,8 @@ func start(_ context.Context, cmd *cli.Command) error {
 		RequestTimeoutGetPayload: time.Duration(cmd.Int(timeoutGetPayloadFlag.Name)) * time.Millisecond,
 		RequestTimeoutRegVal:     time.Duration(cmd.Int(timeoutRegValFlag.Name)) * time.Millisecond,
 		RequestMaxRetries:        int(cmd.Int(maxRetriesFlag.Name)),
+		PrometheusListenAddr:     int(cmd.Int(prometheusListenAddr.Name)),
+		PrometheusRegistry:       prometheusRegistry,
 	}
 	service, err := server.NewBoostService(opts)
 	if err != nil {
@@ -93,6 +105,15 @@ func start(_ context.Context, cmd *cli.Command) error {
 		log.Error("no relay passed the health-check!")
 	}
 
+	if opts.PrometheusListenAddr > 0 && opts.PrometheusListenAddr <= math.MaxUint16 {
+		go func() {
+			log.Infof("Metric Server Listening on %d", opts.PrometheusListenAddr)
+			if err := service.StartMetricsServer(); err != nil {
+				log.WithError(err).Error("metrics server exited with error")
+			}
+		}()
+	}
+
 	log.Infof("Listening on %v", listenAddr)
 	return service.StartHTTPServer()
 }

go.mod

@@ -9,6 +9,7 @@ require (
 	github.com/google/uuid v1.6.0
 	github.com/gorilla/mux v1.8.1
 	github.com/holiman/uint256 v1.3.1
+	github.com/prometheus/client_golang v1.16.0
 	github.com/prysmaticlabs/go-bitfield v0.0.0-20240328144219-a1caa50c3a1e
 	github.com/sirupsen/logrus v1.9.3
 	github.com/stretchr/testify v1.9.0
@@ -44,7 +45,6 @@ require (
 	github.com/mattn/go-isatty v0.0.20 // indirect
 	github.com/matttproud/golang_protobuf_extensions v1.0.4 // indirect
 	github.com/mmcloughlin/addchain v0.4.0 // indirect
-	github.com/prometheus/client_golang v1.16.0 // indirect
 	github.com/prometheus/client_model v0.3.0 // indirect
 	github.com/prometheus/common v0.42.0 // indirect
 	github.com/prometheus/procfs v0.10.1 // indirect

server/service.go

@@ -27,6 +27,8 @@ import (
 	"github.com/flashbots/mev-boost/server/types"
 	"github.com/google/uuid"
 	"github.com/gorilla/mux"
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
 	"github.com/sirupsen/logrus"
 )
 
@@ -37,6 +39,7 @@ var (
 	errInvalidPubkey             = errors.New("invalid pubkey")
 	errNoSuccessfulRelayResponse = errors.New("no successful relay response")
 	errServerAlreadyRunning      = errors.New("server already running")
+	errNilPrometheusRegistry     = errors.New("nil prometheus registry")
 )
 
 var (
@@ -69,6 +72,8 @@ type BoostServiceOpts struct {
 	RequestTimeoutGetPayload time.Duration
 	RequestTimeoutRegVal     time.Duration
 	RequestMaxRetries        int
+	PrometheusListenAddr     int
+	PrometheusRegistry       *prometheus.Registry
 }
 
 // BoostService - the mev-boost service
@@ -93,6 +98,9 @@ type BoostService struct {
 
 	slotUID     *slotUID
 	slotUIDLock sync.Mutex
+
+	prometheusListenAddr int
+	prometheusRegistry   *prometheus.Registry
 }
 
 // NewBoostService created a new BoostService
@@ -130,7 +138,9 @@ func NewBoostService(opts BoostServiceOpts) (*BoostService, error) {
 			Timeout:       opts.RequestTimeoutRegVal,
 			CheckRedirect: httpClientDisallowRedirects,
 		},
-		requestMaxRetries: opts.RequestMaxRetries,
+		requestMaxRetries:    opts.RequestMaxRetries,
+		prometheusListenAddr: opts.PrometheusListenAddr,
+		prometheusRegistry:   opts.PrometheusRegistry,
 	}, nil
 }
 
@@ -194,6 +204,22 @@ func (m *BoostService) StartHTTPServer() error {
 	return err
 }
 
+// StartMetricsServer starts the HTTP server for exporting open-metrics
+func (m *BoostService) StartMetricsServer() error {
+	if m.prometheusRegistry == nil {
+		return errNilPrometheusRegistry
+	}
+	serveMux := http.NewServeMux()
+	serveMux.Handle("/metrics", promhttp.HandlerFor(m.prometheusRegistry, promhttp.HandlerOpts{
+		ErrorLog:          m.log,
+		EnableOpenMetrics: true,
+	}))
+	return http.ListenAndServe(
+		fmt.Sprintf(":%d", m.prometheusListenAddr),
+		serveMux,
+	)
+}
+
 func (m *BoostService) startBidCacheCleanupTask() {
 	for {
 		time.Sleep(1 * time.Minute)