feat: otel tracing

api/repo/create.go

@@ -98,7 +98,7 @@ func CreateRepo(c *gin.Context) {
 	}).Infof("creating new repo %s", input.GetFullName())
 
 	// get repo information from the source
-	r, err := scm.FromContext(c).GetRepo(u, input)
+	r, err := scm.FromContext(c).GetRepo(c.Request.Context(), u, input)
 	if err != nil {
 		retErr := fmt.Errorf("unable to retrieve repo info for %s from source: %w", r.GetFullName(), err)
 
@@ -256,7 +256,7 @@ func CreateRepo(c *gin.Context) {
 	// check if we should create the webhook
 	if c.Value("webhookvalidation").(bool) {
 		// send API call to create the webhook
-		h, _, err = scm.FromContext(c).Enable(u, r, h)
+		h, _, err = scm.FromContext(c).Enable(c.Request.Context(), u, r, h)
 		if err != nil {
 			retErr := fmt.Errorf("unable to create webhook for %s: %w", r.GetFullName(), err)
 

api/repo/delete.go

@@ -70,7 +70,7 @@ func DeleteRepo(c *gin.Context) {
 	}).Infof("deleting repo %s", r.GetFullName())
 
 	// send API call to remove the webhook
-	err := scm.FromContext(c).Disable(u, r.GetOrg(), r.GetName())
+	err := scm.FromContext(c).Disable(c.Request.Context(), u, r.GetOrg(), r.GetName())
 	if err != nil {
 		retErr := fmt.Errorf("unable to delete webhook for %s: %w", r.GetFullName(), err)
 

api/repo/repair.go

@@ -68,7 +68,7 @@ func RepairRepo(c *gin.Context) {
 	// check if we should create the webhook
 	if c.Value("webhookvalidation").(bool) {
 		// send API call to remove the webhook
-		err := scm.FromContext(c).Disable(u, r.GetOrg(), r.GetName())
+		err := scm.FromContext(c).Disable(c.Request.Context(), u, r.GetOrg(), r.GetName())
 		if err != nil {
 			retErr := fmt.Errorf("unable to delete webhook for %s: %w", r.GetFullName(), err)
 
@@ -87,7 +87,7 @@ func RepairRepo(c *gin.Context) {
 		}
 
 		// send API call to create the webhook
-		hook, _, err = scm.FromContext(c).Enable(u, r, hook)
+		hook, _, err = scm.FromContext(c).Enable(c.Request.Context(), u, r, hook)
 		if err != nil {
 			retErr := fmt.Errorf("unable to create webhook for %s: %w", r.GetFullName(), err)
 

api/schedule/create.go

@@ -183,7 +183,7 @@ func CreateSchedule(c *gin.Context) {
 		s, _ = database.FromContext(c).GetScheduleForRepo(r, dbSchedule.GetName())
 	} else {
 		// send API call to create the schedule
-		err = database.FromContext(c).CreateSchedule(s)
+		err = database.FromContext(c).CreateSchedule(c.Request.Context(), s)
 		if err != nil {
 			retErr := fmt.Errorf("unable to create new schedule %s: %w", r.GetName(), err)
 

api/scm/sync.go

@@ -72,7 +72,7 @@ func SyncRepo(c *gin.Context) {
 	logger.Infof("syncing repo %s", r.GetFullName())
 
 	// retrieve repo from source code manager service
-	_, err := scm.FromContext(c).GetRepo(u, r)
+	_, err := scm.FromContext(c).GetRepo(c.Request.Context(), u, r)
 
 	// if there is an error retrieving repo, we know it is deleted: set to inactive
 	if err != nil {

api/scm/sync_org.go

@@ -108,7 +108,7 @@ func SyncReposForOrg(c *gin.Context) {
 
 	// iterate through captured repos and check if they are in GitHub
 	for _, repo := range repos {
-		_, err := scm.FromContext(c).GetRepo(u, repo)
+		_, err := scm.FromContext(c).GetRepo(c.Request.Context(), u, repo)
 		// if repo cannot be captured from GitHub, set to inactive in database
 		if err != nil {
 			repo.SetActive(false)

api/user/get_source.go

@@ -6,6 +6,7 @@ package user
 
 import (
 	"fmt"
+	"go.opentelemetry.io/otel/trace"
 	"net/http"
 
 	"github.com/gin-gonic/gin"
@@ -41,20 +42,23 @@ import (
 func GetSourceRepos(c *gin.Context) {
 	// capture middleware values
 	u := user.Retrieve(c)
+	ctx := c.Request.Context()
 
 	// update engine logger with API metadata
 	//
 	// https://pkg.go.dev/github.com/sirupsen/logrus?tab=doc#Entry.WithFields
 	logrus.WithFields(logrus.Fields{
-		"user": u.GetName(),
+		"user":     u.GetName(),
+		"span_id":  trace.SpanFromContext(ctx).SpanContext().SpanID(),
+		"trace_id": trace.SpanFromContext(ctx).SpanContext().TraceID(),
 	}).Infof("reading available SCM repos for user %s", u.GetName())
 
 	// variables to capture requested data
 	dbRepos := []*library.Repo{}
 	output := make(map[string][]library.Repo)
 
 	// send API call to capture the list of repos for the user
-	srcRepos, err := scm.FromContext(c).ListUserRepos(u)
+	srcRepos, err := scm.FromContext(c).ListUserRepos(ctx, u)
 	if err != nil {
 		retErr := fmt.Errorf("unable to get SCM repos for user %s: %w", u.GetName(), err)
 

cmd/vela-server/main.go

@@ -16,6 +16,7 @@ import (
 	"github.com/go-vela/server/queue"
 	"github.com/go-vela/server/scm"
 	"github.com/go-vela/server/secret"
+	"github.com/go-vela/server/tracing"
 	"github.com/go-vela/server/version"
 	"github.com/sirupsen/logrus"
 	"github.com/urfave/cli/v2"
@@ -235,6 +236,9 @@ func main() {
 	// Add Source Flags
 	app.Flags = append(app.Flags, scm.Flags...)
 
+	// Add Tracing Flags
+	app.Flags = append(app.Flags, tracing.Flags...)
+
 	// set logrus to log in JSON format
 	logrus.SetFormatter(&logrus.JSONFormatter{})
 

cmd/vela-server/server.go

@@ -18,8 +18,10 @@ import (
 	"github.com/go-vela/server/database"
 	"github.com/go-vela/server/router"
 	"github.com/go-vela/server/router/middleware"
+	"github.com/go-vela/server/tracing"
 	"github.com/sirupsen/logrus"
 	"github.com/urfave/cli/v2"
+	"go.opentelemetry.io/contrib/instrumentation/github.com/gin-gonic/gin/otelgin"
 	"golang.org/x/sync/errgroup"
 
 	"k8s.io/apimachinery/pkg/util/wait"
@@ -62,7 +64,29 @@ func server(c *cli.Context) error {
 		return err
 	}
 
-	database, err := database.FromCLIContext(c)
+	// create the tracing config
+	tc, err := tracing.New(c)
+	if err != nil {
+		return err
+	}
+
+	// opt-in for telemetry
+	tracingMiddleware := func(ctx *gin.Context) {}
+	if tc.EnableTracing {
+		// initialize tracing middleware
+		tracingMiddleware = otelgin.Middleware(c.String("tracing.service.name"),
+			otelgin.WithTracerProvider(tc.TracerProvider))
+
+		// shutdown tracing when server closes
+		defer func() {
+			err := tc.TracerProvider.Shutdown(context.Background())
+			if err != nil {
+				logrus.Errorf("unable to shutdown tracer provider: %v", err)
+			}
+		}()
+	}
+
+	database, err := database.FromCLIContext(c, tc)
 	if err != nil {
 		return err
 	}
@@ -108,6 +132,9 @@ func server(c *cli.Context) error {
 		middleware.DefaultRepoEvents(c.StringSlice("default-repo-events")),
 		middleware.AllowlistSchedule(c.StringSlice("vela-schedule-allowlist")),
 		middleware.ScheduleFrequency(c.Duration("schedule-minimum-frequency")),
+
+		// tracing middle, which is an empty handler when tracing is disabled
+		tracingMiddleware,
 	)
 
 	addr, err := url.Parse(c.String("server-addr"))

database/context.go

@@ -7,6 +7,7 @@ package database
 import (
 	"context"
 
+	"github.com/go-vela/server/tracing"
 	"github.com/sirupsen/logrus"
 	"github.com/urfave/cli/v2"
 )
@@ -30,6 +31,8 @@ func FromContext(c context.Context) Interface {
 		return nil
 	}
 
+	// assign as inflight property
+
 	return d
 }
 
@@ -40,7 +43,7 @@ func ToContext(c Setter, d Interface) {
 }
 
 // FromCLIContext creates and returns a database engine from the urfave/cli context.
-func FromCLIContext(c *cli.Context) (Interface, error) {
+func FromCLIContext(c *cli.Context, tc *tracing.TracingConfig) (Interface, error) {
 	logrus.Debug("creating database engine from CLI configuration")
 
 	return New(
@@ -52,5 +55,6 @@ func FromCLIContext(c *cli.Context) (Interface, error) {
 		WithDriver(c.String("database.driver")),
 		WithEncryptionKey(c.String("database.encryption.key")),
 		WithSkipCreation(c.Bool("database.skip_creation")),
+		WithTracingConfig(tc),
 	)
 }

database/context_test.go

@@ -134,7 +134,7 @@ func TestDatabase_FromCLIContext(t *testing.T) {
 	// run tests
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			_, err := FromCLIContext(test.context)
+			_, err := FromCLIContext(test.context, nil)
 
 			if test.failure {
 				if err == nil {

database/database.go

@@ -19,8 +19,10 @@ import (
 	"github.com/go-vela/server/database/step"
 	"github.com/go-vela/server/database/user"
 	"github.com/go-vela/server/database/worker"
+	"github.com/go-vela/server/tracing"
 	"github.com/go-vela/types/constants"
 	"github.com/sirupsen/logrus"
+	"github.com/uptrace/opentelemetry-go-extra/otelgorm"
 
 	"gorm.io/driver/postgres"
 	"gorm.io/driver/sqlite"
@@ -56,6 +58,8 @@ type (
 		config *config
 		// sirupsen/logrus logger used in database functions
 		logger *logrus.Entry
+		// configurations related to telemetry/tracing
+		tracing *tracing.TracingConfig
 
 		build.BuildInterface
 		hook.HookInterface
@@ -129,6 +133,19 @@ func New(opts ...EngineOpt) (Interface, error) {
 		return nil, err
 	}
 
+	// initialize otel tracing if enabled
+	if e.tracing.EnableTracing {
+		otelPlugin := otelgorm.NewPlugin(
+			otelgorm.WithTracerProvider(e.tracing.TracerProvider),
+			otelgorm.WithoutQueryVariables(),
+		)
+
+		err := e.client.Use(otelPlugin)
+		if err != nil {
+			return nil, err
+		}
+	}
+
 	// set the maximum amount of time a connection may be reused
 	db.SetConnMaxLifetime(e.config.ConnectionLife)
 	// set the maximum number of connections in the idle connection pool

database/opts.go

@@ -4,7 +4,11 @@
 
 package database
 
-import "time"
+import (
+	"time"
+
+	"github.com/go-vela/server/tracing"
+)
 
 // EngineOpt represents a configuration option to initialize the database engine.
 type EngineOpt func(*engine) error
@@ -88,3 +92,13 @@ func WithSkipCreation(skipCreation bool) EngineOpt {
 		return nil
 	}
 }
+
+// WithTracingConfig sets the shared tracing config in the database engine.
+func WithTracingConfig(tracing *tracing.TracingConfig) EngineOpt {
+	return func(e *engine) error {
+		// set the tracing config
+		e.tracing = tracing
+
+		return nil
+	}
+}

database/schedule/count_active_test.go

@@ -5,6 +5,7 @@
 package schedule
 
 import (
+	"context"
 	"reflect"
 	"testing"
 
@@ -46,12 +47,14 @@ func TestSchedule_Engine_CountActiveSchedules(t *testing.T) {
 	_sqlite := testSqlite(t)
 	defer func() { _sql, _ := _sqlite.client.DB(); _sql.Close() }()
 
-	err := _sqlite.CreateSchedule(_scheduleOne)
+	ctx := context.TODO()
+
+	err := _sqlite.CreateSchedule(ctx, _scheduleOne)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}
 
-	err = _sqlite.CreateSchedule(_scheduleTwo)
+	err = _sqlite.CreateSchedule(ctx, _scheduleTwo)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}

database/schedule/count_repo_test.go

@@ -5,6 +5,7 @@
 package schedule
 
 import (
+	"context"
 	"reflect"
 	"testing"
 
@@ -50,12 +51,14 @@ func TestSchedule_Engine_CountSchedulesForRepo(t *testing.T) {
 	_sqlite := testSqlite(t)
 	defer func() { _sql, _ := _sqlite.client.DB(); _sql.Close() }()
 
-	err := _sqlite.CreateSchedule(_scheduleOne)
+	ctx := context.TODO()
+
+	err := _sqlite.CreateSchedule(ctx, _scheduleOne)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}
 
-	err = _sqlite.CreateSchedule(_scheduleTwo)
+	err = _sqlite.CreateSchedule(ctx, _scheduleTwo)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}

database/schedule/count_test.go

@@ -5,6 +5,7 @@
 package schedule
 
 import (
+	"context"
 	"reflect"
 	"testing"
 
@@ -44,12 +45,14 @@ func TestSchedule_Engine_CountSchedules(t *testing.T) {
 	_sqlite := testSqlite(t)
 	defer func() { _sql, _ := _sqlite.client.DB(); _sql.Close() }()
 
-	err := _sqlite.CreateSchedule(_scheduleOne)
+	ctx := context.TODO()
+
+	err := _sqlite.CreateSchedule(ctx, _scheduleOne)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}
 
-	err = _sqlite.CreateSchedule(_scheduleTwo)
+	err = _sqlite.CreateSchedule(ctx, _scheduleTwo)
 	if err != nil {
 		t.Errorf("unable to create test schedule for sqlite: %v", err)
 	}

database/schedule/create.go

@@ -6,16 +6,21 @@
 package schedule
 
 import (
+	"context"
+
 	"github.com/go-vela/types/constants"
 	"github.com/go-vela/types/database"
 	"github.com/go-vela/types/library"
 	"github.com/sirupsen/logrus"
+	"go.opentelemetry.io/otel/trace"
 )
 
 // CreateSchedule creates a new schedule in the database.
-func (e *engine) CreateSchedule(s *library.Schedule) error {
+func (e *engine) CreateSchedule(ctx context.Context, s *library.Schedule) error {
 	e.logger.WithFields(logrus.Fields{
 		"schedule": s.GetName(),
+		"span_id":  trace.SpanFromContext(ctx).SpanContext().SpanID(),
+		"trace_id": trace.SpanFromContext(ctx).SpanContext().TraceID(),
 	}).Tracef("creating schedule %s in the database", s.GetName())
 
 	// cast the library type to database type
@@ -29,6 +34,7 @@ func (e *engine) CreateSchedule(s *library.Schedule) error {
 
 	// send query to the database
 	return e.client.
+		WithContext(ctx).
 		Table(constants.TableSchedule).
 		Create(schedule).
 		Error