image_pull.go

/*
   Copyright The containerd Authors.

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.
*/

package sbserver

import (
	"context"
	"crypto/tls"
	"crypto/x509"
	"encoding/base64"
	"fmt"
	"io"
	"net"
	"net/http"
	"net/url"
	"os"
	"path/filepath"
	"strconv"
	"strings"
	"sync"
	"sync/atomic"
	"time"

	"github.com/containerd/imgcrypt"
	"github.com/containerd/imgcrypt/images/encryption"
	imagespec "github.com/opencontainers/image-spec/specs-go/v1"
	runtime "k8s.io/cri-api/pkg/apis/runtime/v1"

	"github.com/containerd/containerd"
	"github.com/containerd/containerd/errdefs"
	containerdimages "github.com/containerd/containerd/images"
	"github.com/containerd/containerd/log"
	"github.com/containerd/containerd/pkg/cri/annotations"
	criconfig "github.com/containerd/containerd/pkg/cri/config"
	crilabels "github.com/containerd/containerd/pkg/cri/labels"
	snpkg "github.com/containerd/containerd/pkg/snapshotters"
	distribution "github.com/containerd/containerd/reference/docker"
	"github.com/containerd/containerd/remotes/docker"
	"github.com/containerd/containerd/remotes/docker/config"
	"github.com/containerd/containerd/tracing"
)

// For image management:
// 1) We have an in-memory metadata index to:
//   a. Maintain ImageID -> RepoTags, ImageID -> RepoDigset relationships; ImageID
//   is the digest of image config, which conforms to oci image spec.
//   b. Cache constant and useful information such as image chainID, config etc.
//   c. An image will be added into the in-memory metadata only when it's successfully
//   pulled and unpacked.
//
// 2) We use containerd image metadata store and content store:
//   a. To resolve image reference (digest/tag) locally. During pulling image, we
//   normalize the image reference provided by user, and put it into image metadata
//   store with resolved descriptor. For the other operations, if image id is provided,
//   we'll access the in-memory metadata index directly; if image reference is
//   provided, we'll normalize it, resolve it in containerd image metadata store
//   to get the image id.
//   b. As the backup of in-memory metadata in 1). During startup, the in-memory
//   metadata could be re-constructed from image metadata store + content store.
//
// Several problems with current approach:
// 1) An entry in containerd image metadata store doesn't mean a "READY" (successfully
// pulled and unpacked) image. E.g. during pulling, the client gets killed. In that case,
// if we saw an image without snapshots or with in-complete contents during startup,
// should we re-pull the image? Or should we remove the entry?
//
// yanxuean: We can't delete image directly, because we don't know if the image
// is pulled by us. There are resource leakage.
//
// 2) Containerd suggests user to add entry before pulling the image. However if
// an error occurs during the pulling, should we remove the entry from metadata
// store? Or should we leave it there until next startup (resource leakage)?
//
// 3) The cri plugin only exposes "READY" (successfully pulled and unpacked) images
// to the user, which are maintained in the in-memory metadata index. However, it's
// still possible that someone else removes the content or snapshot by-pass the cri plugin,
// how do we detect that and update the in-memory metadata correspondingly? Always
// check whether corresponding snapshot is ready when reporting image status?
//
// 4) Is the content important if we cached necessary information in-memory
// after we pull the image? How to manage the disk usage of contents? If some
// contents are missing but snapshots are ready, is the image still "READY"?

// PullImage pulls an image with authentication config.
func (c *criService) PullImage(ctx context.Context, r *runtime.PullImageRequest) (_ *runtime.PullImageResponse, err error) {
	span := tracing.SpanFromContext(ctx)
	defer func() {
		// TODO: add domain label for imagePulls metrics, and we may need to provide a mechanism
		// for the user to configure the set of registries that they are interested in.
		if err != nil {
			imagePulls.WithValues("failure").Inc()
		} else {
			imagePulls.WithValues("success").Inc()
		}
	}()

	inProgressImagePulls.Inc()
	defer inProgressImagePulls.Dec()
	startTime := time.Now()

	imageRef := r.GetImage().GetImage()
	namedRef, err := distribution.ParseDockerRef(imageRef)
	if err != nil {
		return nil, fmt.Errorf("failed to parse image reference %q: %w", imageRef, err)
	}
	ref := namedRef.String()
	if ref != imageRef {
		log.G(ctx).Debugf("PullImage using normalized image ref: %q", ref)
	}

	imagePullProgressTimeout, err := time.ParseDuration(c.config.ImagePullProgressTimeout)
	if err != nil {
		return nil, fmt.Errorf("failed to parse image_pull_progress_timeout %q: %w", c.config.ImagePullProgressTimeout, err)
	}

	var (
		pctx, pcancel = context.WithCancel(ctx)

		pullReporter = newPullProgressReporter(ref, pcancel, imagePullProgressTimeout)

		resolver = docker.NewResolver(docker.ResolverOptions{
			Headers: c.config.Registry.Headers,
			Hosts:   c.registryHosts(ctx, r.GetAuth(), pullReporter.optionUpdateClient),
		})
		isSchema1    bool
		imageHandler containerdimages.HandlerFunc = func(_ context.Context,
			desc imagespec.Descriptor) ([]imagespec.Descriptor, error) {
			if desc.MediaType == containerdimages.MediaTypeDockerSchema1Manifest {
				isSchema1 = true
			}
			return nil, nil
		}
	)

	defer pcancel()
	snapshotter, err := c.snapshotterFromPodSandboxConfig(ctx, ref, r.SandboxConfig)
	if err != nil {
		return nil, err
	}
	log.G(ctx).Debugf("PullImage %q with snapshotter %s", ref, snapshotter)
	span.SetAttributes(
		tracing.Attribute("image.ref", ref),
		tracing.Attribute("snapshotter.name", snapshotter),
	)

	labels := c.getLabels(ctx, ref)

	pullOpts := []containerd.RemoteOpt{
		containerd.WithSchema1Conversion, //nolint:staticcheck // Ignore SA1019. Need to keep deprecated package for compatibility.
		containerd.WithResolver(resolver),
		containerd.WithPullSnapshotter(snapshotter),
		containerd.WithPullUnpack,
		containerd.WithPullLabels(labels),
		containerd.WithMaxConcurrentDownloads(c.config.MaxConcurrentDownloads),
		containerd.WithImageHandler(imageHandler),
		containerd.WithUnpackOpts([]containerd.UnpackOpt{
			containerd.WithUnpackDuplicationSuppressor(c.unpackDuplicationSuppressor),
		}),
	}

	pullOpts = append(pullOpts, c.encryptedImagesPullOpts()...)
	if !c.config.ContainerdConfig.DisableSnapshotAnnotations {
		pullOpts = append(pullOpts,
			containerd.WithImageHandlerWrapper(snpkg.AppendInfoHandlerWrapper(ref)))
	}

	if c.config.ContainerdConfig.DiscardUnpackedLayers {
		// Allows GC to clean layers up from the content store after unpacking
		pullOpts = append(pullOpts,
			containerd.WithChildLabelMap(containerdimages.ChildGCLabelsFilterLayers))
	}

	pullReporter.start(pctx)
	image, err := c.client.Pull(pctx, ref, pullOpts...)
	pcancel()
	if err != nil {
		return nil, fmt.Errorf("failed to pull and unpack image %q: %w", ref, err)
	}
	span.AddEvent("Pull and unpack image complete")

	configDesc, err := image.Config(ctx)
	if err != nil {
		return nil, fmt.Errorf("get image config descriptor: %w", err)
	}
	imageID := configDesc.Digest.String()

	repoDigest, repoTag := getRepoDigestAndTag(namedRef, image.Target().Digest, isSchema1)
	for _, r := range []string{imageID, repoTag, repoDigest} {
		if r == "" {
			continue
		}
		if err := c.createImageReference(ctx, r, image.Target(), labels); err != nil {
			return nil, fmt.Errorf("failed to create image reference %q: %w", r, err)
		}
		// Update image store to reflect the newest state in containerd.
		// No need to use `updateImage`, because the image reference must
		// have been managed by the cri plugin.
		if err := c.imageStore.Update(ctx, r); err != nil {
			return nil, fmt.Errorf("failed to update image store %q: %w", r, err)
		}
	}

	const mbToByte = 1024 * 1024
	size, _ := image.Size(ctx)
	imagePullingSpeed := float64(size) / mbToByte / time.Since(startTime).Seconds()
	imagePullThroughput.Observe(imagePullingSpeed)

	log.G(ctx).Infof("Pulled image %q with image id %q, repo tag %q, repo digest %q, size %q in %s", imageRef, imageID,
		repoTag, repoDigest, strconv.FormatInt(size, 10), time.Since(startTime))
	// NOTE(random-liu): the actual state in containerd is the source of truth, even we maintain
	// in-memory image store, it's only for in-memory indexing. The image could be removed
	// by someone else anytime, before/during/after we create the metadata. We should always
	// check the actual state in containerd before using the image or returning status of the
	// image.
	return &runtime.PullImageResponse{ImageRef: imageID}, nil
}

// ParseAuth parses AuthConfig and returns username and password/secret required by containerd.
func ParseAuth(auth *runtime.AuthConfig, host string) (string, string, error) {
	if auth == nil {
		return "", "", nil
	}
	if auth.ServerAddress != "" {
		// Do not return the auth info when server address doesn't match.
		u, err := url.Parse(auth.ServerAddress)
		if err != nil {
			return "", "", fmt.Errorf("parse server address: %w", err)
		}
		if host != u.Host {
			return "", "", nil
		}
	}
	if auth.Username != "" {
		return auth.Username, auth.Password, nil
	}
	if auth.IdentityToken != "" {
		return "", auth.IdentityToken, nil
	}
	if auth.Auth != "" {
		decLen := base64.StdEncoding.DecodedLen(len(auth.Auth))
		decoded := make([]byte, decLen)
		_, err := base64.StdEncoding.Decode(decoded, []byte(auth.Auth))
		if err != nil {
			return "", "", err
		}
		user, passwd, ok := strings.Cut(string(decoded), ":")
		if !ok {
			return "", "", fmt.Errorf("invalid decoded auth: %q", decoded)
		}
		return user, strings.Trim(passwd, "\x00"), nil
	}
	// TODO(random-liu): Support RegistryToken.
	// An empty auth config is valid for anonymous registry
	return "", "", nil
}

// createImageReference creates image reference inside containerd image store.
// Note that because create and update are not finished in one transaction, there could be race. E.g.
// the image reference is deleted by someone else after create returns already exists, but before update
// happens.
func (c *criService) createImageReference(ctx context.Context, name string, desc imagespec.Descriptor, labels map[string]string) error {
	img := containerdimages.Image{
		Name:   name,
		Target: desc,
		// Add a label to indicate that the image is managed by the cri plugin.
		Labels: labels,
	}
	// TODO(random-liu): Figure out which is the more performant sequence create then update or
	// update then create.
	_, err := c.client.ImageService().Create(ctx, img)
	if err == nil || !errdefs.IsAlreadyExists(err) {
		return err
	}
	// Retrieve oldImg from image store here because Create routine returns an
	// empty image on ErrAlreadyExists
	oldImg, err := c.client.ImageService().Get(ctx, name)
	if err != nil {
		return err
	}
	fieldpaths := []string{"target"}
	if oldImg.Labels[crilabels.ImageLabelKey] != labels[crilabels.ImageLabelKey] {
		fieldpaths = append(fieldpaths, "labels."+crilabels.ImageLabelKey)
	}
	if oldImg.Labels[crilabels.PinnedImageLabelKey] != labels[crilabels.PinnedImageLabelKey] &&
		labels[crilabels.PinnedImageLabelKey] == crilabels.PinnedImageLabelValue {
		fieldpaths = append(fieldpaths, "labels."+crilabels.PinnedImageLabelKey)
	}
	if oldImg.Target.Digest == img.Target.Digest && len(fieldpaths) < 2 {
		return nil
	}
	_, err = c.client.ImageService().Update(ctx, img, fieldpaths...)
	return err
}

// getLabels get image labels to be added on CRI image
func (c *criService) getLabels(ctx context.Context, name string) map[string]string {
	labels := map[string]string{crilabels.ImageLabelKey: crilabels.ImageLabelValue}
	configSandboxImage := c.config.SandboxImage
	// parse sandbox image
	sandboxNamedRef, err := distribution.ParseDockerRef(configSandboxImage)
	if err != nil {
		log.G(ctx).Errorf("failed to parse sandbox image from config %s", sandboxNamedRef)
		return nil
	}
	sandboxRef := sandboxNamedRef.String()
	// Adding pinned image label to sandbox image
	if sandboxRef == name {
		labels[crilabels.PinnedImageLabelKey] = crilabels.PinnedImageLabelValue
	}
	return labels
}

// updateImage updates image store to reflect the newest state of an image reference
// in containerd. If the reference is not managed by the cri plugin, the function also
// generates necessary metadata for the image and make it managed.
func (c *criService) updateImage(ctx context.Context, r string) error {
	img, err := c.client.GetImage(ctx, r)
	if err != nil && !errdefs.IsNotFound(err) {
		return fmt.Errorf("get image by reference: %w", err)
	}
	if err == nil && img.Labels()[crilabels.ImageLabelKey] != crilabels.ImageLabelValue {
		// Make sure the image has the image id as its unique
		// identifier that references the image in its lifetime.
		configDesc, err := img.Config(ctx)
		if err != nil {
			return fmt.Errorf("get image id: %w", err)
		}
		id := configDesc.Digest.String()
		labels := c.getLabels(ctx, r)
		if err := c.createImageReference(ctx, id, img.Target(), labels); err != nil {
			return fmt.Errorf("create image id reference %q: %w", id, err)
		}
		if err := c.imageStore.Update(ctx, id); err != nil {
			return fmt.Errorf("update image store for %q: %w", id, err)
		}
		// The image id is ready, add the label to mark the image as managed.
		if err := c.createImageReference(ctx, r, img.Target(), labels); err != nil {
			return fmt.Errorf("create managed label: %w", err)
		}
	}
	// If the image is not found, we should continue updating the cache,
	// so that the image can be removed from the cache.
	if err := c.imageStore.Update(ctx, r); err != nil {
		return fmt.Errorf("update image store for %q: %w", r, err)
	}
	return nil
}

// getTLSConfig returns a TLSConfig configured with a CA/Cert/Key specified by registryTLSConfig
func (c *criService) getTLSConfig(registryTLSConfig criconfig.TLSConfig) (*tls.Config, error) {
	var (
		tlsConfig = &tls.Config{}
		cert      tls.Certificate
		err       error
	)
	if registryTLSConfig.CertFile != "" && registryTLSConfig.KeyFile == "" {
		return nil, fmt.Errorf("cert file %q was specified, but no corresponding key file was specified", registryTLSConfig.CertFile)
	}
	if registryTLSConfig.CertFile == "" && registryTLSConfig.KeyFile != "" {
		return nil, fmt.Errorf("key file %q was specified, but no corresponding cert file was specified", registryTLSConfig.KeyFile)
	}
	if registryTLSConfig.CertFile != "" && registryTLSConfig.KeyFile != "" {
		cert, err = tls.LoadX509KeyPair(registryTLSConfig.CertFile, registryTLSConfig.KeyFile)
		if err != nil {
			return nil, fmt.Errorf("failed to load cert file: %w", err)
		}
		if len(cert.Certificate) != 0 {
			tlsConfig.Certificates = []tls.Certificate{cert}
		}
		// TODO(thaJeztah): verify if we should ignore the deprecation; see https://github.com/containerd/containerd/pull/7349/files#r990644833
		tlsConfig.BuildNameToCertificate() //nolint:staticcheck
	}

	if registryTLSConfig.CAFile != "" {
		caCertPool, err := x509.SystemCertPool()
		if err != nil {
			return nil, fmt.Errorf("failed to get system cert pool: %w", err)
		}
		caCert, err := os.ReadFile(registryTLSConfig.CAFile)
		if err != nil {
			return nil, fmt.Errorf("failed to load CA file: %w", err)
		}
		caCertPool.AppendCertsFromPEM(caCert)
		tlsConfig.RootCAs = caCertPool
	}

	tlsConfig.InsecureSkipVerify = registryTLSConfig.InsecureSkipVerify
	return tlsConfig, nil
}

func hostDirFromRoots(roots []string) func(string) (string, error) {
	rootfn := make([]func(string) (string, error), len(roots))
	for i := range roots {
		rootfn[i] = config.HostDirFromRoot(roots[i])
	}
	return func(host string) (dir string, err error) {
		for _, fn := range rootfn {
			dir, err = fn(host)
			if (err != nil && !errdefs.IsNotFound(err)) || (dir != "") {
				break
			}
		}
		return
	}
}

// registryHosts is the registry hosts to be used by the resolver.
func (c *criService) registryHosts(ctx context.Context, auth *runtime.AuthConfig, updateClientFn config.UpdateClientFunc) docker.RegistryHosts {
	paths := filepath.SplitList(c.config.Registry.ConfigPath)
	if len(paths) > 0 {
		hostOptions := config.HostOptions{
			UpdateClient: updateClientFn,
		}
		hostOptions.Credentials = func(host string) (string, string, error) {
			hostauth := auth
			if hostauth == nil {
				config := c.config.Registry.Configs[host]
				if config.Auth != nil {
					hostauth = toRuntimeAuthConfig(*config.Auth)
				}
			}
			return ParseAuth(hostauth, host)
		}
		hostOptions.HostDir = hostDirFromRoots(paths)

		return config.ConfigureHosts(ctx, hostOptions)
	}

	return func(host string) ([]docker.RegistryHost, error) {
		var registries []docker.RegistryHost

		endpoints, err := c.registryEndpoints(host)
		if err != nil {
			return nil, fmt.Errorf("get registry endpoints: %w", err)
		}
		for _, e := range endpoints {
			u, err := url.Parse(e)
			if err != nil {
				return nil, fmt.Errorf("parse registry endpoint %q from mirrors: %w", e, err)
			}

			var (
				transport = newTransport()
				client    = &http.Client{Transport: transport}
				config    = c.config.Registry.Configs[u.Host]
			)

			if config.TLS != nil {
				transport.TLSClientConfig, err = c.getTLSConfig(*config.TLS)
				if err != nil {
					return nil, fmt.Errorf("get TLSConfig for registry %q: %w", e, err)
				}
			} else if docker.IsLocalhost(host) && u.Scheme == "http" {
				// Skipping TLS verification for localhost
				transport.TLSClientConfig = &tls.Config{
					InsecureSkipVerify: true,
				}
			}

			// Make a copy of `auth`, so that different authorizers would not reference
			// the same auth variable.
			auth := auth
			if auth == nil && config.Auth != nil {
				auth = toRuntimeAuthConfig(*config.Auth)
			}

			if updateClientFn != nil {
				if err := updateClientFn(client); err != nil {
					return nil, fmt.Errorf("failed to update http client: %w", err)
				}
			}

			authorizer := docker.NewDockerAuthorizer(
				docker.WithAuthClient(client),
				docker.WithAuthCreds(func(host string) (string, string, error) {
					return ParseAuth(auth, host)
				}))

			if u.Path == "" {
				u.Path = "/v2"
			}

			registries = append(registries, docker.RegistryHost{
				Client:       client,
				Authorizer:   authorizer,
				Host:         u.Host,
				Scheme:       u.Scheme,
				Path:         u.Path,
				Capabilities: docker.HostCapabilityResolve | docker.HostCapabilityPull,
			})
		}
		return registries, nil
	}
}

// defaultScheme returns the default scheme for a registry host.
func defaultScheme(host string) string {
	if docker.IsLocalhost(host) {
		return "http"
	}
	return "https"
}

// addDefaultScheme returns the endpoint with default scheme
func addDefaultScheme(endpoint string) (string, error) {
	if strings.Contains(endpoint, "://") {
		return endpoint, nil
	}
	ue := "dummy://" + endpoint
	u, err := url.Parse(ue)
	if err != nil {
		return "", err
	}
	return fmt.Sprintf("%s://%s", defaultScheme(u.Host), endpoint), nil
}

// registryEndpoints returns endpoints for a given host.
// It adds default registry endpoint if it does not exist in the passed-in endpoint list.
// It also supports wildcard host matching with `*`.
func (c *criService) registryEndpoints(host string) ([]string, error) {
	var endpoints []string
	_, ok := c.config.Registry.Mirrors[host]
	if ok {
		endpoints = c.config.Registry.Mirrors[host].Endpoints
	} else {
		endpoints = c.config.Registry.Mirrors["*"].Endpoints
	}
	defaultHost, err := docker.DefaultHost(host)
	if err != nil {
		return nil, fmt.Errorf("get default host: %w", err)
	}
	for i := range endpoints {
		en, err := addDefaultScheme(endpoints[i])
		if err != nil {
			return nil, fmt.Errorf("parse endpoint url: %w", err)
		}
		endpoints[i] = en
	}
	for _, e := range endpoints {
		u, err := url.Parse(e)
		if err != nil {
			return nil, fmt.Errorf("parse endpoint url: %w", err)
		}
		if u.Host == host {
			// Do not add default if the endpoint already exists.
			return endpoints, nil
		}
	}
	return append(endpoints, defaultScheme(defaultHost)+"://"+defaultHost), nil
}

// newTransport returns a new HTTP transport used to pull image.
// TODO(random-liu): Create a library and share this code with `ctr`.
func newTransport() *http.Transport {
	return &http.Transport{
		Proxy: http.ProxyFromEnvironment,
		DialContext: (&net.Dialer{
			Timeout:       30 * time.Second,
			KeepAlive:     30 * time.Second,
			FallbackDelay: 300 * time.Millisecond,
		}).DialContext,
		MaxIdleConns:          10,
		IdleConnTimeout:       30 * time.Second,
		TLSHandshakeTimeout:   10 * time.Second,
		ExpectContinueTimeout: 5 * time.Second,
	}
}

// encryptedImagesPullOpts returns the necessary list of pull options required
// for decryption of encrypted images based on the cri decryption configuration.
func (c *criService) encryptedImagesPullOpts() []containerd.RemoteOpt {
	if c.config.ImageDecryption.KeyModel == criconfig.KeyModelNode {
		ltdd := imgcrypt.Payload{}
		decUnpackOpt := encryption.WithUnpackConfigApplyOpts(encryption.WithDecryptedUnpack(&ltdd))
		opt := containerd.WithUnpackOpts([]containerd.UnpackOpt{decUnpackOpt})
		return []containerd.RemoteOpt{opt}
	}
	return nil
}

const (
	// minPullProgressReportInternal is used to prevent the reporter from
	// eating more CPU resources
	minPullProgressReportInternal = 5 * time.Second
	// defaultPullProgressReportInterval represents that how often the
	// reporter checks that pull progress.
	defaultPullProgressReportInterval = 10 * time.Second
)

// pullProgressReporter is used to check single PullImage progress.
type pullProgressReporter struct {
	ref         string
	cancel      context.CancelFunc
	reqReporter pullRequestReporter
	timeout     time.Duration
}

func newPullProgressReporter(ref string, cancel context.CancelFunc, timeout time.Duration) *pullProgressReporter {
	return &pullProgressReporter{
		ref:         ref,
		cancel:      cancel,
		reqReporter: pullRequestReporter{},
		timeout:     timeout,
	}
}

func (reporter *pullProgressReporter) optionUpdateClient(client *http.Client) error {
	client.Transport = &pullRequestReporterRoundTripper{
		rt:          client.Transport,
		reqReporter: &reporter.reqReporter,
	}
	return nil
}

func (reporter *pullProgressReporter) start(ctx context.Context) {
	if reporter.timeout == 0 {
		log.G(ctx).Infof("no timeout and will not start pulling image %s reporter", reporter.ref)
		return
	}

	go func() {
		var (
			reportInterval = defaultPullProgressReportInterval

			lastSeenBytesRead = uint64(0)
			lastSeenTimestamp = time.Now()
		)

		// check progress more frequently if timeout < default internal
		if reporter.timeout < reportInterval {
			reportInterval = reporter.timeout / 2

			if reportInterval < minPullProgressReportInternal {
				reportInterval = minPullProgressReportInternal
			}
		}

		var ticker = time.NewTicker(reportInterval)
		defer ticker.Stop()

		for {
			select {
			case <-ticker.C:
				activeReqs, bytesRead := reporter.reqReporter.status()

				log.G(ctx).WithField("ref", reporter.ref).
					WithField("activeReqs", activeReqs).
					WithField("totalBytesRead", bytesRead).
					WithField("lastSeenBytesRead", lastSeenBytesRead).
					WithField("lastSeenTimestamp", lastSeenTimestamp).
					WithField("reportInterval", reportInterval).
					Tracef("progress for image pull")

				if activeReqs == 0 || bytesRead > lastSeenBytesRead {
					lastSeenBytesRead = bytesRead
					lastSeenTimestamp = time.Now()
					continue
				}

				if time.Since(lastSeenTimestamp) > reporter.timeout {
					log.G(ctx).Errorf("cancel pulling image %s because of no progress in %v", reporter.ref, reporter.timeout)
					reporter.cancel()
					return
				}
			case <-ctx.Done():
				activeReqs, bytesRead := reporter.reqReporter.status()
				log.G(ctx).Infof("stop pulling image %s: active requests=%v, bytes read=%v", reporter.ref, activeReqs, bytesRead)
				return
			}
		}
	}()
}

// countingReadCloser wraps http.Response.Body with pull request reporter,
// which is used by pullRequestReporterRoundTripper.
type countingReadCloser struct {
	once sync.Once

	rc          io.ReadCloser
	reqReporter *pullRequestReporter
}

// Read reads bytes from original io.ReadCloser and increases bytes in
// pull request reporter.
func (r *countingReadCloser) Read(p []byte) (int, error) {
	n, err := r.rc.Read(p)
	r.reqReporter.incByteRead(uint64(n))
	return n, err
}

// Close closes the original io.ReadCloser and only decreases the number of
// active pull requests once.
func (r *countingReadCloser) Close() error {
	err := r.rc.Close()
	r.once.Do(r.reqReporter.decRequest)
	return err
}

// pullRequestReporter is used to track the progress per each criapi.PullImage.
type pullRequestReporter struct {
	// activeReqs indicates that current number of active pulling requests,
	// including auth requests.
	activeReqs int32
	// totalBytesRead indicates that the total bytes has been read from
	// remote registry.
	totalBytesRead uint64
}

func (reporter *pullRequestReporter) incRequest() {
	atomic.AddInt32(&reporter.activeReqs, 1)
}

func (reporter *pullRequestReporter) decRequest() {
	atomic.AddInt32(&reporter.activeReqs, -1)
}

func (reporter *pullRequestReporter) incByteRead(nr uint64) {
	atomic.AddUint64(&reporter.totalBytesRead, nr)
}

func (reporter *pullRequestReporter) status() (currentReqs int32, totalBytesRead uint64) {
	currentReqs = atomic.LoadInt32(&reporter.activeReqs)
	totalBytesRead = atomic.LoadUint64(&reporter.totalBytesRead)
	return currentReqs, totalBytesRead
}

// pullRequestReporterRoundTripper wraps http.RoundTripper with pull request
// reporter which is used to track the progress of active http request with
// counting readable http.Response.Body.
//
// NOTE:
//
// Although containerd provides ingester manager to track the progress
// of pulling request, for example `ctr image pull` shows the console progress
// bar, it needs more CPU resources to open/read the ingested files with
// acquiring containerd metadata plugin's boltdb lock.
//
// Before sending HTTP request to registry, the containerd.Client.Pull library
// will open writer by containerd ingester manager. Based on this, the
// http.RoundTripper wrapper can track the active progress with lower overhead
// even if the ref has been locked in ingester manager by other Pull request.
type pullRequestReporterRoundTripper struct {
	rt http.RoundTripper

	reqReporter *pullRequestReporter
}

func (rt *pullRequestReporterRoundTripper) RoundTrip(req *http.Request) (*http.Response, error) {
	rt.reqReporter.incRequest()

	resp, err := rt.rt.RoundTrip(req)
	if err != nil {
		rt.reqReporter.decRequest()
		return nil, err
	}

	resp.Body = &countingReadCloser{
		rc:          resp.Body,
		reqReporter: rt.reqReporter,
	}
	return resp, err
}

// Given that runtime information is not passed from PullImageRequest, we depend on an experimental annotation
// passed from pod sandbox config to get the runtimeHandler. The annotation key is specified in configuration.
// Once we know the runtime, try to override default snapshotter if it is set for this runtime.
// See https://github.com/containerd/containerd/issues/6657
func (c *criService) snapshotterFromPodSandboxConfig(ctx context.Context, imageRef string,
	s *runtime.PodSandboxConfig) (string, error) {
	snapshotter := c.config.ContainerdConfig.Snapshotter
	if s == nil || s.Annotations == nil {
		return snapshotter, nil
	}

	runtimeHandler, ok := s.Annotations[annotations.RuntimeHandler]
	if !ok {
		return snapshotter, nil
	}

	ociRuntime, err := c.getSandboxRuntime(s, runtimeHandler)
	if err != nil {
		return "", fmt.Errorf("experimental: failed to get sandbox runtime for %s, err: %+v", runtimeHandler, err)
	}

	snapshotter = c.runtimeSnapshotter(context.Background(), ociRuntime)
	log.G(ctx).Infof("experimental: PullImage %q for runtime %s, using snapshotter %s", imageRef, runtimeHandler, snapshotter)
	return snapshotter, nil
}