moby/daemon/prune.go

260 lines
7.2 KiB
Go
Raw Permalink Normal View History

package daemon // import "github.com/docker/docker/daemon"
import (
"context"
"regexp"
API: add "prune" events This patch adds a new "prune" event type to indicate that pruning of a resource type completed. This event-type can be used on systems that want to perform actions after resources have been cleaned up. For example, Docker Desktop performs an fstrim after resources are deleted (https://github.com/linuxkit/linuxkit/tree/v0.7/pkg/trim-after-delete). While the current (remove, destroy) events can provide information on _most_ resources, there is currently no event triggered after the BuildKit build-cache is cleaned. Prune events have a `reclaimed` attribute, indicating the amount of space that was reclaimed (in bytes). The attribute can be used, for example, to use as a threshold for performing fstrim actions. Reclaimed space for `network` events will always be 0, but the field is added to be consistent with prune events for other resources. To test this patch: Create some resources: for i in foo bar baz; do \ docker network create network_$i \ && docker volume create volume_$i \ && docker run -d --name container_$i -v volume_$i:/volume busybox sh -c 'truncate -s 5M somefile; truncate -s 5M /volume/file' \ && docker tag busybox:latest image_$i; \ done; docker pull alpine docker pull nginx:alpine echo -e "FROM busybox\nRUN truncate -s 50M bigfile" | DOCKER_BUILDKIT=1 docker build - Start listening for "prune" events in another shell: docker events --filter event=prune Prune containers, networks, volumes, and build-cache: docker system prune -af --volumes See the events that are returned: docker events --filter event=prune 2020-07-25T12:12:09.268491000Z container prune (reclaimed=15728640) 2020-07-25T12:12:09.447890400Z network prune (reclaimed=0) 2020-07-25T12:12:09.452323000Z volume prune (reclaimed=15728640) 2020-07-25T12:12:09.517236200Z image prune (reclaimed=21568540) 2020-07-25T12:12:09.566662600Z builder prune (reclaimed=52428841) Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2020-07-25 12:14:38 +00:00
"strconv"
"sync/atomic"
"time"
"github.com/containerd/log"
"github.com/docker/docker/api/types"
"github.com/docker/docker/api/types/backend"
API: add "prune" events This patch adds a new "prune" event type to indicate that pruning of a resource type completed. This event-type can be used on systems that want to perform actions after resources have been cleaned up. For example, Docker Desktop performs an fstrim after resources are deleted (https://github.com/linuxkit/linuxkit/tree/v0.7/pkg/trim-after-delete). While the current (remove, destroy) events can provide information on _most_ resources, there is currently no event triggered after the BuildKit build-cache is cleaned. Prune events have a `reclaimed` attribute, indicating the amount of space that was reclaimed (in bytes). The attribute can be used, for example, to use as a threshold for performing fstrim actions. Reclaimed space for `network` events will always be 0, but the field is added to be consistent with prune events for other resources. To test this patch: Create some resources: for i in foo bar baz; do \ docker network create network_$i \ && docker volume create volume_$i \ && docker run -d --name container_$i -v volume_$i:/volume busybox sh -c 'truncate -s 5M somefile; truncate -s 5M /volume/file' \ && docker tag busybox:latest image_$i; \ done; docker pull alpine docker pull nginx:alpine echo -e "FROM busybox\nRUN truncate -s 50M bigfile" | DOCKER_BUILDKIT=1 docker build - Start listening for "prune" events in another shell: docker events --filter event=prune Prune containers, networks, volumes, and build-cache: docker system prune -af --volumes See the events that are returned: docker events --filter event=prune 2020-07-25T12:12:09.268491000Z container prune (reclaimed=15728640) 2020-07-25T12:12:09.447890400Z network prune (reclaimed=0) 2020-07-25T12:12:09.452323000Z volume prune (reclaimed=15728640) 2020-07-25T12:12:09.517236200Z image prune (reclaimed=21568540) 2020-07-25T12:12:09.566662600Z builder prune (reclaimed=52428841) Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2020-07-25 12:14:38 +00:00
"github.com/docker/docker/api/types/events"
"github.com/docker/docker/api/types/filters"
timetypes "github.com/docker/docker/api/types/time"
"github.com/docker/docker/errdefs"
"github.com/docker/docker/libnetwork"
"github.com/docker/docker/runconfig"
"github.com/pkg/errors"
)
var (
// errPruneRunning is returned when a prune request is received while
// one is in progress
errPruneRunning = errdefs.Conflict(errors.New("a prune operation is already running"))
containersAcceptedFilters = map[string]bool{
"label": true,
"label!": true,
"until": true,
}
networksAcceptedFilters = map[string]bool{
"label": true,
"label!": true,
"until": true,
}
)
// ContainersPrune removes unused containers
func (daemon *Daemon) ContainersPrune(ctx context.Context, pruneFilters filters.Args) (*types.ContainersPruneReport, error) {
if !atomic.CompareAndSwapInt32(&daemon.pruneRunning, 0, 1) {
return nil, errPruneRunning
}
defer atomic.StoreInt32(&daemon.pruneRunning, 0)
rep := &types.ContainersPruneReport{}
// make sure that only accepted filters have been received
err := pruneFilters.Validate(containersAcceptedFilters)
if err != nil {
return nil, err
}
until, err := getUntilFromPruneFilters(pruneFilters)
if err != nil {
return nil, err
}
daemon: reload runtimes w/o breaking containers The existing runtimes reload logic went to great lengths to replace the directory containing runtime wrapper scripts as atomically as possible within the limitations of the Linux filesystem ABI. Trouble is, atomically swapping the wrapper scripts directory solves the wrong problem! The runtime configuration is "locked in" when a container is started, including the path to the runC binary. If a container is started with a runtime which requires a daemon-managed wrapper script and then the daemon is reloaded with a config which no longer requires the wrapper script (i.e. some args -> no args, or the runtime is dropped from the config), that container would become unmanageable. Any attempts to stop, exec or otherwise perform lifecycle management operations on the container are likely to fail due to the wrapper script no longer existing at its original path. Atomically swapping the wrapper scripts is also incompatible with the read-copy-update paradigm for reloading configuration. A handler in the daemon could retain a reference to the pre-reload configuration for an indeterminate amount of time after the daemon configuration has been reloaded and updated. It is possible for the daemon to attempt to start a container using a deleted wrapper script if a request to run a container races a reload. Solve the problem of deleting referenced wrapper scripts by ensuring that all wrapper scripts are *immutable* for the lifetime of the daemon process. Any given runtime wrapper script must always exist with the same contents, no matter how many times the daemon config is reloaded, or what changes are made to the config. This is accomplished by using everyone's favourite design pattern: content-addressable storage. Each wrapper script file name is suffixed with the SHA-256 digest of its contents to (probabilistically) guarantee immutability without needing any concurrency control. Stale runtime wrapper scripts are only cleaned up on the next daemon restart. Split the derived runtimes configuration from the user-supplied configuration to have a place to store derived state without mutating the user-supplied configuration or exposing daemon internals in API struct types. Hold the derived state and the user-supplied configuration in a single struct value so that they can be updated as an atomic unit. Signed-off-by: Cory Snider <csnider@mirantis.com>
2022-08-31 20:12:30 +00:00
cfg := &daemon.config().Config
allContainers := daemon.List()
for _, c := range allContainers {
select {
case <-ctx.Done():
log.G(ctx).Debugf("ContainersPrune operation cancelled: %#v", *rep)
return rep, nil
default:
}
if !c.IsRunning() {
if !until.IsZero() && c.Created.After(until) {
continue
}
if !matchLabels(pruneFilters, c.Config.Labels) {
continue
}
cSize, _, err := daemon.imageService.GetContainerLayerSize(ctx, c.ID)
if err != nil {
return nil, err
}
// TODO: sets RmLink to true?
err = daemon.containerRm(cfg, c.ID, &backend.ContainerRmConfig{})
if err != nil {
log.G(ctx).Warnf("failed to prune container %s: %v", c.ID, err)
continue
}
if cSize > 0 {
rep.SpaceReclaimed += uint64(cSize)
}
rep.ContainersDeleted = append(rep.ContainersDeleted, c.ID)
}
}
daemon.EventsService.Log(events.ActionPrune, events.ContainerEventType, events.Actor{
API: add "prune" events This patch adds a new "prune" event type to indicate that pruning of a resource type completed. This event-type can be used on systems that want to perform actions after resources have been cleaned up. For example, Docker Desktop performs an fstrim after resources are deleted (https://github.com/linuxkit/linuxkit/tree/v0.7/pkg/trim-after-delete). While the current (remove, destroy) events can provide information on _most_ resources, there is currently no event triggered after the BuildKit build-cache is cleaned. Prune events have a `reclaimed` attribute, indicating the amount of space that was reclaimed (in bytes). The attribute can be used, for example, to use as a threshold for performing fstrim actions. Reclaimed space for `network` events will always be 0, but the field is added to be consistent with prune events for other resources. To test this patch: Create some resources: for i in foo bar baz; do \ docker network create network_$i \ && docker volume create volume_$i \ && docker run -d --name container_$i -v volume_$i:/volume busybox sh -c 'truncate -s 5M somefile; truncate -s 5M /volume/file' \ && docker tag busybox:latest image_$i; \ done; docker pull alpine docker pull nginx:alpine echo -e "FROM busybox\nRUN truncate -s 50M bigfile" | DOCKER_BUILDKIT=1 docker build - Start listening for "prune" events in another shell: docker events --filter event=prune Prune containers, networks, volumes, and build-cache: docker system prune -af --volumes See the events that are returned: docker events --filter event=prune 2020-07-25T12:12:09.268491000Z container prune (reclaimed=15728640) 2020-07-25T12:12:09.447890400Z network prune (reclaimed=0) 2020-07-25T12:12:09.452323000Z volume prune (reclaimed=15728640) 2020-07-25T12:12:09.517236200Z image prune (reclaimed=21568540) 2020-07-25T12:12:09.566662600Z builder prune (reclaimed=52428841) Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2020-07-25 12:14:38 +00:00
Attributes: map[string]string{"reclaimed": strconv.FormatUint(rep.SpaceReclaimed, 10)},
})
return rep, nil
}
// localNetworksPrune removes unused local networks
func (daemon *Daemon) localNetworksPrune(ctx context.Context, pruneFilters filters.Args) *types.NetworksPruneReport {
rep := &types.NetworksPruneReport{}
until, _ := getUntilFromPruneFilters(pruneFilters)
// When the function returns true, the walk will stop.
daemon.netController.WalkNetworks(func(nw *libnetwork.Network) bool {
select {
case <-ctx.Done():
// context cancelled
return true
default:
}
if nw.ConfigOnly() {
return false
}
if !until.IsZero() && nw.Created().After(until) {
return false
}
if !matchLabels(pruneFilters, nw.Labels()) {
return false
}
nwName := nw.Name()
if runconfig.IsPreDefinedNetwork(nwName) {
return false
}
if len(nw.Endpoints()) > 0 {
return false
}
if err := daemon.DeleteNetwork(nw.ID()); err != nil {
log.G(ctx).Warnf("could not remove local network %s: %v", nwName, err)
return false
}
rep.NetworksDeleted = append(rep.NetworksDeleted, nwName)
return false
})
return rep
}
// clusterNetworksPrune removes unused cluster networks
func (daemon *Daemon) clusterNetworksPrune(ctx context.Context, pruneFilters filters.Args) (*types.NetworksPruneReport, error) {
rep := &types.NetworksPruneReport{}
until, _ := getUntilFromPruneFilters(pruneFilters)
cluster := daemon.GetCluster()
if !cluster.IsManager() {
return rep, nil
}
networks, err := cluster.GetNetworks(pruneFilters)
if err != nil {
return rep, err
}
networkIsInUse := regexp.MustCompile(`network ([[:alnum:]]+) is in use`)
for _, nw := range networks {
select {
case <-ctx.Done():
return rep, nil
default:
if nw.Ingress {
// Routing-mesh network removal has to be explicitly invoked by user
continue
}
if !until.IsZero() && nw.Created.After(until) {
continue
}
if !matchLabels(pruneFilters, nw.Labels) {
continue
}
// https://github.com/docker/docker/issues/24186
// `docker network inspect` unfortunately displays ONLY those containers that are local to that node.
// So we try to remove it anyway and check the error
err = cluster.RemoveNetwork(nw.ID)
if err != nil {
// we can safely ignore the "network .. is in use" error
match := networkIsInUse.FindStringSubmatch(err.Error())
if len(match) != 2 || match[1] != nw.ID {
log.G(ctx).Warnf("could not remove cluster network %s: %v", nw.Name, err)
}
continue
}
rep.NetworksDeleted = append(rep.NetworksDeleted, nw.Name)
}
}
return rep, nil
}
// NetworksPrune removes unused networks
func (daemon *Daemon) NetworksPrune(ctx context.Context, pruneFilters filters.Args) (*types.NetworksPruneReport, error) {
if !atomic.CompareAndSwapInt32(&daemon.pruneRunning, 0, 1) {
return nil, errPruneRunning
}
defer atomic.StoreInt32(&daemon.pruneRunning, 0)
// make sure that only accepted filters have been received
err := pruneFilters.Validate(networksAcceptedFilters)
if err != nil {
return nil, err
}
if _, err := getUntilFromPruneFilters(pruneFilters); err != nil {
return nil, err
}
rep := &types.NetworksPruneReport{}
if clusterRep, err := daemon.clusterNetworksPrune(ctx, pruneFilters); err == nil {
rep.NetworksDeleted = append(rep.NetworksDeleted, clusterRep.NetworksDeleted...)
}
localRep := daemon.localNetworksPrune(ctx, pruneFilters)
rep.NetworksDeleted = append(rep.NetworksDeleted, localRep.NetworksDeleted...)
select {
case <-ctx.Done():
log.G(ctx).Debugf("NetworksPrune operation cancelled: %#v", *rep)
return rep, nil
default:
}
daemon.EventsService.Log(events.ActionPrune, events.NetworkEventType, events.Actor{
API: add "prune" events This patch adds a new "prune" event type to indicate that pruning of a resource type completed. This event-type can be used on systems that want to perform actions after resources have been cleaned up. For example, Docker Desktop performs an fstrim after resources are deleted (https://github.com/linuxkit/linuxkit/tree/v0.7/pkg/trim-after-delete). While the current (remove, destroy) events can provide information on _most_ resources, there is currently no event triggered after the BuildKit build-cache is cleaned. Prune events have a `reclaimed` attribute, indicating the amount of space that was reclaimed (in bytes). The attribute can be used, for example, to use as a threshold for performing fstrim actions. Reclaimed space for `network` events will always be 0, but the field is added to be consistent with prune events for other resources. To test this patch: Create some resources: for i in foo bar baz; do \ docker network create network_$i \ && docker volume create volume_$i \ && docker run -d --name container_$i -v volume_$i:/volume busybox sh -c 'truncate -s 5M somefile; truncate -s 5M /volume/file' \ && docker tag busybox:latest image_$i; \ done; docker pull alpine docker pull nginx:alpine echo -e "FROM busybox\nRUN truncate -s 50M bigfile" | DOCKER_BUILDKIT=1 docker build - Start listening for "prune" events in another shell: docker events --filter event=prune Prune containers, networks, volumes, and build-cache: docker system prune -af --volumes See the events that are returned: docker events --filter event=prune 2020-07-25T12:12:09.268491000Z container prune (reclaimed=15728640) 2020-07-25T12:12:09.447890400Z network prune (reclaimed=0) 2020-07-25T12:12:09.452323000Z volume prune (reclaimed=15728640) 2020-07-25T12:12:09.517236200Z image prune (reclaimed=21568540) 2020-07-25T12:12:09.566662600Z builder prune (reclaimed=52428841) Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2020-07-25 12:14:38 +00:00
Attributes: map[string]string{"reclaimed": "0"},
})
return rep, nil
}
func getUntilFromPruneFilters(pruneFilters filters.Args) (time.Time, error) {
until := time.Time{}
if !pruneFilters.Contains("until") {
return until, nil
}
untilFilters := pruneFilters.Get("until")
if len(untilFilters) > 1 {
return until, errdefs.InvalidParameter(errors.New("more than one until filter specified"))
}
ts, err := timetypes.GetTimestamp(untilFilters[0], time.Now())
if err != nil {
return until, errdefs.InvalidParameter(err)
}
seconds, nanoseconds, err := timetypes.ParseTimestamps(ts, 0)
if err != nil {
return until, errdefs.InvalidParameter(err)
}
until = time.Unix(seconds, nanoseconds)
return until, nil
}
func matchLabels(pruneFilters filters.Args, labels map[string]string) bool {
if !pruneFilters.MatchKVList("label", labels) {
return false
}
// By default MatchKVList will return true if field (like 'label!') does not exist
// So we have to add additional Contains("label!") check
if pruneFilters.Contains("label!") {
if pruneFilters.MatchKVList("label!", labels) {
return false
}
}
return true
}