Add otel tracing support

Signed-off-by: Brian Goff <cpuguy83@gmail.com>

Author: cpuguy83

cli/command/container/stats.go

@@ -19,6 +19,8 @@ import (
 	"github.com/docker/docker/api/types/filters"
 	"github.com/pkg/errors"
 	"github.com/spf13/cobra"
+	"go.opentelemetry.io/otel/attribute"
+	"go.opentelemetry.io/otel/trace"
 )
 
 type statsOptions struct {
@@ -125,13 +127,16 @@ func runStats(ctx context.Context, dockerCli command.Cli, opts *statsOptions) er
 	}
 
 	if showAll {
+		span := trace.SpanFromContext(ctx)
+
 		// If no names were specified, start a long running goroutine which
 		// monitors container events. We make sure we're subscribed before
 		// retrieving the list of running containers to avoid a race where we
 		// would "miss" a creation.
 		started := make(chan struct{})
 		eh := command.InitEventHandler()
 		eh.Handle(events.ActionCreate, func(e events.Message) {
+			span.AddEvent(string(events.ActionCreate), trace.WithAttributes(attribute.String("ID", e.ID)))
 			if opts.all {
 				s := NewStats(e.Actor.ID[:12])
 				if cStats.add(s) {
@@ -142,6 +147,7 @@ func runStats(ctx context.Context, dockerCli command.Cli, opts *statsOptions) er
 		})
 
 		eh.Handle(events.ActionStart, func(e events.Message) {
+			span.AddEvent(string(events.ActionStart), trace.WithAttributes(attribute.String("ID", e.Actor.ID)))
 			s := NewStats(e.Actor.ID[:12])
 			if cStats.add(s) {
 				waitFirst.Add(1)
@@ -150,6 +156,7 @@ func runStats(ctx context.Context, dockerCli command.Cli, opts *statsOptions) er
 		})
 
 		eh.Handle(events.ActionDie, func(e events.Message) {
+			span.AddEvent(string(events.ActionDie), trace.WithAttributes(attribute.String("ID", e.ID)))
 			if !opts.all {
 				cStats.remove(e.Actor.ID[:12])
 			}

cli/command/manifest/rm.go

@@ -8,6 +8,7 @@ import (
 	"github.com/docker/cli/cli/command"
 	"github.com/pkg/errors"
 	"github.com/spf13/cobra"
+	"go.opentelemetry.io/otel/trace"
 )
 
 func newRmManifestListCommand(dockerCli command.Cli) *cobra.Command {
@@ -23,11 +24,14 @@ func newRmManifestListCommand(dockerCli command.Cli) *cobra.Command {
 	return cmd
 }
 
-func runRm(_ context.Context, dockerCli command.Cli, targets []string) error {
+func runRm(ctx context.Context, dockerCli command.Cli, targets []string) error {
+	span := trace.SpanFromContext(ctx)
+
 	var errs []string
 	for _, target := range targets {
 		targetRef, refErr := normalizeReference(target)
 		if refErr != nil {
+			span.RecordError(refErr)
 			errs = append(errs, refErr.Error())
 			continue
 		}

cli/command/registry/logout.go

@@ -8,6 +8,7 @@ import (
 	"github.com/docker/cli/cli/command"
 	"github.com/docker/docker/registry"
 	"github.com/spf13/cobra"
+	"go.opentelemetry.io/otel/trace"
 )
 
 // NewLogoutCommand creates a new `docker logout` command
@@ -33,7 +34,7 @@ func NewLogoutCommand(dockerCli command.Cli) *cobra.Command {
 	return cmd
 }
 
-func runLogout(_ context.Context, dockerCli command.Cli, serverAddress string) error {
+func runLogout(ctx context.Context, dockerCli command.Cli, serverAddress string) error {
 	var isDefaultRegistry bool
 
 	if serverAddress == "" {
@@ -52,10 +53,13 @@ func runLogout(_ context.Context, dockerCli command.Cli, serverAddress string) e
 		regsToLogout = append(regsToLogout, hostnameAddress, "http://"+hostnameAddress, "https://"+hostnameAddress)
 	}
 
+	span := trace.SpanFromContext(ctx)
+
 	fmt.Fprintf(dockerCli.Out(), "Removing login credentials for %s\n", hostnameAddress)
 	errs := make(map[string]error)
 	for _, r := range regsToLogout {
 		if err := dockerCli.ConfigFile().GetCredentialsStore(r).Erase(r); err != nil {
+			span.RecordError(err)
 			errs[r] = err
 		}
 	}

cmd/docker/docker.go

@@ -1,6 +1,7 @@
 package main
 
 import (
+	"context"
 	"fmt"
 	"os"
 	"os/exec"
@@ -15,10 +16,14 @@ import (
 	"github.com/docker/cli/cli/version"
 	"github.com/docker/cli/cmd/docker/internal/appcontext"
 	"github.com/docker/docker/api/types/versions"
+	"github.com/moby/buildkit/util/tracing/detect"
 	"github.com/pkg/errors"
 	"github.com/sirupsen/logrus"
 	"github.com/spf13/cobra"
 	"github.com/spf13/pflag"
+	"go.opentelemetry.io/otel"
+	"go.opentelemetry.io/otel/propagation"
+	"go.opentelemetry.io/otel/trace"
 )
 
 func newDockerCommand(dockerCli *command.DockerCli) *cli.TopLevelCommand {
@@ -40,7 +45,25 @@ func newDockerCommand(dockerCli *command.DockerCli) *cli.TopLevelCommand {
 			return fmt.Errorf("docker: '%s' is not a docker command.\nSee 'docker --help'", args[0])
 		},
 		PersistentPreRunE: func(cmd *cobra.Command, args []string) error {
-			return isSupported(cmd, dockerCli)
+			if err := isSupported(cmd, dockerCli); err != nil {
+				return err
+			}
+
+			name := cmd.Name()
+			for p := cmd.Parent(); p != nil && p != cmd.Root(); p = p.Parent() {
+				name = p.Name() + " " + name
+			}
+
+			ctx, _ := otel.Tracer("").Start(cmd.Context(), name)
+			cmd.SetContext(ctx)
+			dockerCli.WithContext(ctx)
+
+			return nil
+		},
+		PersistentPostRun: func(cmd *cobra.Command, args []string) {
+			// TODO: There doesn't seem to be a way to determine if the command returned an an error
+			// so we can set the span status here.
+			trace.SpanFromContext(cmd.Context()).End()
 		},
 		Version:               fmt.Sprintf("%s, build %s", version.Version, version.GitCommit),
 		DisableFlagsInUseLine: true,
@@ -54,6 +77,10 @@ func newDockerCommand(dockerCli *command.DockerCli) *cli.TopLevelCommand {
 	cmd.SetOut(dockerCli.Out())
 	cmd.SetErr(dockerCli.Err())
 
+	// Cobra's context may be nil in some cases, so initialize it here.
+	ctx := context.TODO()
+	cmd.SetContext(ctx)
+
 	opts, helpCmd = cli.SetupRootCommand(cmd)
 	registerCompletionFuncForGlobalFlags(dockerCli, cmd)
 	cmd.Flags().BoolP("version", "v", false, "Print version information and quit")
@@ -227,6 +254,29 @@ func runDocker(dockerCli *command.DockerCli) error {
 		return err
 	}
 
+	// Buildkit's detect package currently follows the old otel spec which defaulted to gRPC.
+	// Since the spec changed to default to http/protobuf.
+	// If these env vars are not set then we set them to the new default so detect will give us the expected protocol.
+	// This is the same as on the dockerd side.
+	// This can be removed after buildkit's detect package is updated.
+	if os.Getenv("OTEL_EXPORTER_OTLP_TRACES_PROTOCOL") == "" && os.Getenv("OTEL_EXPORTER_OTLP_PROTOCOL") == "" {
+		os.Setenv("OTEL_EXPORTER_OTLP_TRACES_PROTOCOL", "http/protobuf")
+	}
+	if v := os.Getenv("OTEL_SERVICE_NAME"); v == "" {
+		os.Setenv("OTEL_SERVICE_NAME", cmd.Root().Name())
+	}
+
+	tp, err := detect.TracerProvider()
+	if err != nil {
+		logrus.WithError(err).Debug("Failed to initialize tracing")
+	}
+
+	otel.SetTextMapPropagator(propagation.NewCompositeTextMapPropagator(propagation.TraceContext{}, propagation.Baggage{}))
+	if tp != nil {
+		otel.SetTracerProvider(tp)
+		defer detect.Shutdown(context.Background())
+	}
+
 	var envs []string
 	args, os.Args, envs, err = processAliases(dockerCli, cmd, args, os.Args)
 	if err != nil {

vendor.mod

@@ -37,6 +37,8 @@ require (
 	github.com/theupdateframework/notary v0.7.1-0.20210315103452-bf96a202a09a
 	github.com/tonistiigi/go-rosetta v0.0.0-20200727161949-f79598599c5d
 	github.com/xeipuuv/gojsonschema v1.2.0
+	go.opentelemetry.io/otel v1.14.0
+	go.opentelemetry.io/otel/trace v1.14.0
 	golang.org/x/sync v0.3.0
 	golang.org/x/sys v0.13.0
 	golang.org/x/term v0.13.0
@@ -51,6 +53,7 @@ require (
 	github.com/Microsoft/go-winio v0.6.1 // indirect
 	github.com/Microsoft/hcsshim v0.11.4 // indirect
 	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/cenkalti/backoff/v4 v4.2.0 // indirect
 	github.com/cespare/xxhash/v2 v2.2.0 // indirect
 	github.com/containerd/log v0.1.0 // indirect
 	github.com/docker/go v1.5.1-1.0.20160303222718-d30aec9fd63c // indirect
@@ -76,9 +79,7 @@ require (
 	github.com/xeipuuv/gojsonreference v0.0.0-20180127040603-bd5ef7bd5415 // indirect
 	go.etcd.io/etcd/raft/v3 v3.5.6 // indirect
 	go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.40.0 // indirect
-	go.opentelemetry.io/otel v1.14.0 // indirect
 	go.opentelemetry.io/otel/metric v0.37.0 // indirect
-	go.opentelemetry.io/otel/trace v1.14.0 // indirect
 	golang.org/x/crypto v0.14.0 // indirect
 	golang.org/x/mod v0.11.0 // indirect
 	golang.org/x/net v0.17.0 // indirect
@@ -88,3 +89,17 @@ require (
 	google.golang.org/grpc v1.58.3 // indirect
 	google.golang.org/protobuf v1.31.0 // indirect
 )
+
+require github.com/moby/buildkit v0.12.4
+
+require (
+	github.com/grpc-ecosystem/grpc-gateway/v2 v2.11.3 // indirect
+	go.opentelemetry.io/otel/exporters/otlp/internal/retry v1.14.0 // indirect
+	go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.14.0 // indirect
+	go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.14.0 // indirect
+	go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.14.0 // indirect
+	go.opentelemetry.io/otel/sdk v1.14.0 // indirect
+	go.opentelemetry.io/proto/otlp v0.19.0 // indirect
+	google.golang.org/genproto v0.0.0-20230711160842-782d3b101e98 // indirect
+	google.golang.org/genproto/googleapis/api v0.0.0-20230711160842-782d3b101e98 // indirect
+)