cri-o/pkg/storage/image.go

486 lines
14 KiB
Go
Raw Normal View History

Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
package storage
import (
"errors"
"fmt"
"net"
"path/filepath"
"regexp"
"strings"
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
"github.com/containers/image/copy"
"github.com/containers/image/docker/reference"
"github.com/containers/image/image"
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
"github.com/containers/image/signature"
istorage "github.com/containers/image/storage"
"github.com/containers/image/transports/alltransports"
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
"github.com/containers/image/types"
"github.com/containers/storage"
distreference "github.com/docker/distribution/reference"
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
)
// ImageResult wraps a subset of information about an image: its ID, its names,
// and the size, if known, or nil if it isn't.
type ImageResult struct {
ID string
Names []string
Size *uint64
}
type indexInfo struct {
name string
secure bool
}
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
type imageService struct {
store storage.Store
defaultTransport string
insecureRegistryCIDRs []*net.IPNet
indexConfigs map[string]*indexInfo
registries []string
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
}
// ImageServer wraps up various CRI-related activities into a reusable
// implementation.
type ImageServer interface {
// ListImages returns list of all images which match the filter.
ListImages(systemContext *types.SystemContext, filter string) ([]ImageResult, error)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
// ImageStatus returns status of an image which matches the filter.
ImageStatus(systemContext *types.SystemContext, filter string) (*ImageResult, error)
// PullImage imports an image from the specified location.
PullImage(systemContext *types.SystemContext, imageName string, options *copy.Options) (types.ImageReference, error)
// RemoveImage deletes the specified image.
RemoveImage(systemContext *types.SystemContext, imageName string) error
// GetStore returns the reference to the storage library Store which
// the image server uses to hold images, and is the destination used
// when it's asked to pull an image.
GetStore() storage.Store
// CanPull preliminary checks whether we're allowed to pull an image
CanPull(imageName string, options *copy.Options) (bool, error)
// ResolveNames takes an image reference and if it's unqualified (w/o hostname),
// it uses crio's default registries to qualify it.
ResolveNames(imageName string) ([]string, error)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
}
func (svc *imageService) getRef(name string) (types.ImageReference, error) {
ref, err := alltransports.ParseImageName(name)
if err != nil {
ref2, err2 := istorage.Transport.ParseStoreReference(svc.store, "@"+name)
if err2 != nil {
ref3, err3 := istorage.Transport.ParseStoreReference(svc.store, name)
if err3 != nil {
return nil, err
}
ref2 = ref3
}
ref = ref2
}
return ref, nil
}
func (svc *imageService) ListImages(systemContext *types.SystemContext, filter string) ([]ImageResult, error) {
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
results := []ImageResult{}
if filter != "" {
ref, err := svc.getRef(filter)
if err != nil {
return nil, err
}
if image, err := istorage.Transport.GetStoreImage(svc.store, ref); err == nil {
img, err := ref.NewImage(systemContext)
if err != nil {
return nil, err
}
size := imageSize(img)
img.Close()
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
results = append(results, ImageResult{
ID: image.ID,
Names: image.Names,
Size: size,
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
})
}
} else {
images, err := svc.store.Images()
if err != nil {
return nil, err
}
for _, image := range images {
ref, err := istorage.Transport.ParseStoreReference(svc.store, "@"+image.ID)
if err != nil {
return nil, err
}
img, err := ref.NewImage(systemContext)
if err != nil {
return nil, err
}
size := imageSize(img)
img.Close()
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
results = append(results, ImageResult{
ID: image.ID,
Names: image.Names,
Size: size,
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
})
}
}
return results, nil
}
func (svc *imageService) ImageStatus(systemContext *types.SystemContext, nameOrID string) (*ImageResult, error) {
ref, err := alltransports.ParseImageName(nameOrID)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if err != nil {
ref2, err2 := istorage.Transport.ParseStoreReference(svc.store, "@"+nameOrID)
if err2 != nil {
ref3, err3 := istorage.Transport.ParseStoreReference(svc.store, nameOrID)
if err3 != nil {
return nil, err
}
ref2 = ref3
}
ref = ref2
}
image, err := istorage.Transport.GetStoreImage(svc.store, ref)
if err != nil {
return nil, err
}
img, err := ref.NewImage(systemContext)
if err != nil {
return nil, err
}
size := imageSize(img)
img.Close()
return &ImageResult{
ID: image.ID,
Names: image.Names,
Size: size,
}, nil
}
func imageSize(img types.Image) *uint64 {
if sum, err := img.Size(); err == nil {
usum := uint64(sum)
return &usum
}
return nil
}
func (svc *imageService) CanPull(imageName string, options *copy.Options) (bool, error) {
srcRef, err := svc.prepareImage(imageName, options)
if err != nil {
return false, err
}
rawSource, err := srcRef.NewImageSource(options.SourceCtx, nil)
if err != nil {
return false, err
}
src, err := image.FromSource(rawSource)
if err != nil {
rawSource.Close()
return false, err
}
src.Close()
return true, nil
}
// prepareImage creates an image reference from an image string and set options
// for the source context
func (svc *imageService) prepareImage(imageName string, options *copy.Options) (types.ImageReference, error) {
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if imageName == "" {
return nil, storage.ErrNotAnImage
}
srcRef, err := alltransports.ParseImageName(imageName)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if err != nil {
if svc.defaultTransport == "" {
return nil, err
}
srcRef2, err2 := alltransports.ParseImageName(svc.defaultTransport + imageName)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if err2 != nil {
return nil, err
}
srcRef = srcRef2
}
if options.SourceCtx == nil {
options.SourceCtx = &types.SystemContext{}
}
hostname := reference.Domain(srcRef.DockerReference())
if secure := svc.isSecureIndex(hostname); !secure {
options.SourceCtx.DockerInsecureSkipTLSVerify = !secure
}
return srcRef, nil
}
func (svc *imageService) PullImage(systemContext *types.SystemContext, imageName string, options *copy.Options) (types.ImageReference, error) {
policy, err := signature.DefaultPolicy(systemContext)
if err != nil {
return nil, err
}
policyContext, err := signature.NewPolicyContext(policy)
if err != nil {
return nil, err
}
if options == nil {
options = &copy.Options{}
}
srcRef, err := svc.prepareImage(imageName, options)
if err != nil {
return nil, err
}
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
dest := imageName
if srcRef.DockerReference() != nil {
dest = srcRef.DockerReference().Name()
if tagged, ok := srcRef.DockerReference().(reference.NamedTagged); ok {
dest = dest + ":" + tagged.Tag()
}
if canonical, ok := srcRef.DockerReference().(reference.Canonical); ok {
dest = dest + "@" + canonical.Digest().String()
}
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
}
destRef, err := istorage.Transport.ParseStoreReference(svc.store, dest)
if err != nil {
return nil, err
}
err = copy.Image(policyContext, destRef, srcRef, options)
if err != nil {
return nil, err
}
return destRef, nil
}
func (svc *imageService) RemoveImage(systemContext *types.SystemContext, nameOrID string) error {
ref, err := alltransports.ParseImageName(nameOrID)
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if err != nil {
ref2, err2 := istorage.Transport.ParseStoreReference(svc.store, "@"+nameOrID)
if err2 != nil {
ref3, err3 := istorage.Transport.ParseStoreReference(svc.store, nameOrID)
if err3 != nil {
return err
}
ref2 = ref3
}
ref = ref2
}
return ref.DeleteImage(systemContext)
}
func (svc *imageService) GetStore() storage.Store {
return svc.store
}
func (svc *imageService) isSecureIndex(indexName string) bool {
if index, ok := svc.indexConfigs[indexName]; ok {
return index.secure
}
host, _, err := net.SplitHostPort(indexName)
if err != nil {
// assume indexName is of the form `host` without the port and go on.
host = indexName
}
addrs, err := net.LookupIP(host)
if err != nil {
ip := net.ParseIP(host)
if ip != nil {
addrs = []net.IP{ip}
}
// if ip == nil, then `host` is neither an IP nor it could be looked up,
// either because the index is unreachable, or because the index is behind an HTTP proxy.
// So, len(addrs) == 0 and we're not aborting.
}
// Try CIDR notation only if addrs has any elements, i.e. if `host`'s IP could be determined.
for _, addr := range addrs {
for _, ipnet := range svc.insecureRegistryCIDRs {
// check if the addr falls in the subnet
if ipnet.Contains(addr) {
return false
}
}
}
return true
}
func isValidHostname(hostname string) bool {
return hostname != "" && !strings.Contains(hostname, "/") &&
(strings.Contains(hostname, ".") ||
strings.Contains(hostname, ":") || hostname == "localhost")
}
func isReferenceFullyQualified(reposName reference.Named) bool {
indexName, _, _ := splitReposName(reposName)
return indexName != ""
}
const (
// defaultHostname is the default built-in hostname
defaultHostname = "docker.io"
// legacyDefaultHostname is automatically converted to DefaultHostname
legacyDefaultHostname = "index.docker.io"
// defaultRepoPrefix is the prefix used for default repositories in default host
defaultRepoPrefix = "library/"
)
// splitReposName breaks a reposName into an index name and remote name
func splitReposName(reposName reference.Named) (indexName string, remoteName reference.Named, err error) {
var remoteNameStr string
indexName, remoteNameStr = distreference.SplitHostname(reposName)
if !isValidHostname(indexName) {
// This is a Docker Index repos (ex: samalba/hipache or ubuntu)
// 'docker.io'
indexName = ""
remoteName = reposName
} else {
remoteName, err = withName(remoteNameStr)
}
return
}
func validateName(name string) error {
if err := validateID(strings.TrimPrefix(name, defaultHostname+"/")); err == nil {
return fmt.Errorf("Invalid repository name (%s), cannot specify 64-byte hexadecimal strings", name)
}
return nil
}
var validHex = regexp.MustCompile(`^([a-f0-9]{64})$`)
// validateID checks whether an ID string is a valid image ID.
func validateID(id string) error {
if ok := validHex.MatchString(id); !ok {
return fmt.Errorf("image ID %q is invalid", id)
}
return nil
}
// withName returns a named object representing the given string. If the input
// is invalid ErrReferenceInvalidFormat will be returned.
func withName(name string) (reference.Named, error) {
name, err := normalize(name)
if err != nil {
return nil, err
}
if err := validateName(name); err != nil {
return nil, err
}
r, err := distreference.WithName(name)
return r, err
}
// splitHostname splits a repository name to hostname and remotename string.
// If no valid hostname is found, empty string will be returned as a resulting
// hostname. Repository name needs to be already validated before.
func splitHostname(name string) (hostname, remoteName string) {
i := strings.IndexRune(name, '/')
if i == -1 || (!strings.ContainsAny(name[:i], ".:") && name[:i] != "localhost") {
hostname, remoteName = "", name
} else {
hostname, remoteName = name[:i], name[i+1:]
}
if hostname == legacyDefaultHostname {
hostname = defaultHostname
}
if hostname == defaultHostname && !strings.ContainsRune(remoteName, '/') {
remoteName = defaultRepoPrefix + remoteName
}
return
}
// normalize returns a repository name in its normalized form, meaning it
// will contain library/ prefix for official images.
func normalize(name string) (string, error) {
host, remoteName := splitHostname(name)
if strings.ToLower(remoteName) != remoteName {
return "", errors.New("invalid reference format: repository name must be lowercase")
}
if host == defaultHostname {
if strings.HasPrefix(remoteName, defaultRepoPrefix) {
remoteName = strings.TrimPrefix(remoteName, defaultRepoPrefix)
}
return host + "/" + remoteName, nil
}
return name, nil
}
func (svc *imageService) ResolveNames(imageName string) ([]string, error) {
r, err := reference.ParseNormalizedNamed(imageName)
if err != nil {
return nil, err
}
if isReferenceFullyQualified(r) {
// this means the image is already fully qualified
return []string{imageName}, nil
}
// we got an unqualified image here, we can't go ahead w/o registries configured
// properly.
if len(svc.registries) == 0 {
return nil, errors.New("no registries configured while trying to pull an unqualified image")
}
// this means we got an image in the form of "busybox"
// we need to use additional registries...
// normalize the unqualified image to be domain/repo/image...
_, rest := splitDomain(r.Name())
images := []string{}
for _, r := range svc.registries {
images = append(images, filepath.Join(r, rest))
}
return images, nil
}
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
// GetImageService returns an ImageServer that uses the passed-in store, and
// which will prepend the passed-in defaultTransport value to an image name if
// a name that's passed to its PullImage() method can't be resolved to an image
// in the store and can't be resolved to a source on its own.
func GetImageService(store storage.Store, defaultTransport string, insecureRegistries []string, registries []string) (ImageServer, error) {
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
if store == nil {
var err error
store, err = storage.GetStore(storage.DefaultStoreOptions)
if err != nil {
return nil, err
}
}
seenRegistries := make(map[string]bool, len(registries))
cleanRegistries := []string{}
for _, r := range registries {
if seenRegistries[r] {
continue
}
cleanRegistries = append(cleanRegistries, r)
seenRegistries[r] = true
}
is := &imageService{
store: store,
defaultTransport: defaultTransport,
indexConfigs: make(map[string]*indexInfo, 0),
insecureRegistryCIDRs: make([]*net.IPNet, 0),
registries: cleanRegistries,
}
insecureRegistries = append(insecureRegistries, "127.0.0.0/8")
// Split --insecure-registry into CIDR and registry-specific settings.
for _, r := range insecureRegistries {
// Check if CIDR was passed to --insecure-registry
_, ipnet, err := net.ParseCIDR(r)
if err == nil {
// Valid CIDR.
is.insecureRegistryCIDRs = append(is.insecureRegistryCIDRs, ipnet)
} else {
// Assume `host:port` if not CIDR.
is.indexConfigs[r] = &indexInfo{
name: r,
secure: false,
}
}
}
return is, nil
Add storage utility functions Add an intermediate API layer that uses containers/storage, and a containers/image that has been patched to use it, to manage images and containers, storing the data that we need to know about containers and pods in the metadata fields provided by containers/storage. While ocid manages pods and containers as different types of items, with disjoint sets of IDs and names, it remains true that every pod includes at least one container. When a container's only purpose is to serve as a home for namespaces that are shared with the other containers in the pod, it is referred to as the pod's infrastructure container. At the storage level, a pod is stored as its set of containers. We keep track of both pod IDs and container IDs in the metadata field of Container objects that the storage library manages for us. Containers which bear the same pod ID are members of the pod which has that ID. Other information about the pod, which ocid needs to remember in order to answer requests for information about the pod, is also kept in the metadata field of its member containers. The container's runtime configuration should be stored in the container's ContainerDirectory, and used as a template. Each time the container is about to be started, its layer should be mounted, that configuration template should be read, the template's rootfs location should be replaced with the mountpoint for the container's layer, and the result should be saved to the container's ContainerRunDirectory, for use as the configuration for the container. Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2016-11-22 18:50:33 +00:00
}