moby--moby/builder/dockerfile/builder.go

317 lines
9.0 KiB
Go
Raw Normal View History

package dockerfile
import (
"bytes"
"fmt"
"io"
"io/ioutil"
"strings"
"github.com/Sirupsen/logrus"
"github.com/docker/docker/api/types"
"github.com/docker/docker/api/types/backend"
"github.com/docker/docker/api/types/container"
"github.com/docker/docker/builder"
"github.com/docker/docker/builder/dockerfile/command"
"github.com/docker/docker/builder/dockerfile/parser"
"github.com/docker/docker/builder/remotecontext"
"github.com/docker/docker/pkg/archive"
"github.com/docker/docker/pkg/chrootarchive"
"github.com/docker/docker/pkg/idtools"
"github.com/docker/docker/pkg/streamformatter"
"github.com/docker/docker/pkg/stringid"
"github.com/pkg/errors"
"golang.org/x/net/context"
"golang.org/x/sync/syncmap"
)
var validCommitCommands = map[string]bool{
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 09:48:13 +00:00
"cmd": true,
"entrypoint": true,
"healthcheck": true,
"env": true,
"expose": true,
"label": true,
"onbuild": true,
"user": true,
"volume": true,
"workdir": true,
}
// BuildManager is shared across all Builder objects
type BuildManager struct {
archiver *archive.Archiver
backend builder.Backend
pathCache pathCache // TODO: make this persistent
}
// NewBuildManager creates a BuildManager
func NewBuildManager(b builder.Backend, idMappings *idtools.IDMappings) *BuildManager {
return &BuildManager{
backend: b,
pathCache: &syncmap.Map{},
archiver: chrootarchive.NewArchiver(idMappings),
}
}
// Build starts a new build from a BuildConfig
func (bm *BuildManager) Build(ctx context.Context, config backend.BuildConfig) (*builder.Result, error) {
buildsTriggered.Inc()
if config.Options.Dockerfile == "" {
config.Options.Dockerfile = builder.DefaultDockerfileName
}
source, dockerfile, err := remotecontext.Detect(config)
if err != nil {
return nil, err
}
if source != nil {
defer func() {
if err := source.Close(); err != nil {
logrus.Debugf("[BUILDER] failed to remove temporary context: %v", err)
}
}()
}
builderOptions := builderOptions{
Options: config.Options,
ProgressWriter: config.ProgressWriter,
Backend: bm.backend,
PathCache: bm.pathCache,
Archiver: bm.archiver,
}
return newBuilder(ctx, builderOptions).build(source, dockerfile)
}
// builderOptions are the dependencies required by the builder
type builderOptions struct {
Options *types.ImageBuildOptions
Backend builder.Backend
ProgressWriter backend.ProgressWriter
PathCache pathCache
Archiver *archive.Archiver
}
// Builder is a Dockerfile builder
// It implements the builder.Backend interface.
type Builder struct {
options *types.ImageBuildOptions
Stdout io.Writer
Stderr io.Writer
Aux *streamformatter.AuxFormatter
Output io.Writer
docker builder.Backend
clientCtx context.Context
archiver *archive.Archiver
buildStages *buildStages
disableCommit bool
buildArgs *buildArgs
imageSources *imageSources
pathCache pathCache
containerManager *containerManager
imageProber ImageProber
}
// newBuilder creates a new Dockerfile builder from an optional dockerfile and a Options.
func newBuilder(clientCtx context.Context, options builderOptions) *Builder {
config := options.Options
if config == nil {
config = new(types.ImageBuildOptions)
}
b := &Builder{
clientCtx: clientCtx,
options: config,
Stdout: options.ProgressWriter.StdoutFormatter,
Stderr: options.ProgressWriter.StderrFormatter,
Aux: options.ProgressWriter.AuxFormatter,
Output: options.ProgressWriter.Output,
docker: options.Backend,
archiver: options.Archiver,
buildArgs: newBuildArgs(config.BuildArgs),
buildStages: newBuildStages(),
imageSources: newImageSources(clientCtx, options),
pathCache: options.PathCache,
imageProber: newImageProber(options.Backend, config.CacheFrom, config.NoCache),
containerManager: newContainerManager(options.Backend),
}
return b
}
// Build runs the Dockerfile builder by parsing the Dockerfile and executing
// the instructions from the file.
func (b *Builder) build(source builder.Source, dockerfile *parser.Result) (*builder.Result, error) {
defer b.imageSources.Unmount()
addNodesForLabelOption(dockerfile.AST, b.options.Labels)
if err := checkDispatchDockerfile(dockerfile.AST); err != nil {
buildsFailed.WithValues(metricsDockerfileSyntaxError).Inc()
return nil, err
}
dispatchState, err := b.dispatchDockerfileWithCancellation(dockerfile, source)
if err != nil {
return nil, err
}
if b.options.Target != "" && !dispatchState.isCurrentStage(b.options.Target) {
buildsFailed.WithValues(metricsBuildTargetNotReachableError).Inc()
return nil, errors.Errorf("failed to reach build target %s in Dockerfile", b.options.Target)
}
b.buildArgs.WarnOnUnusedBuildArgs(b.Stderr)
if dispatchState.imageID == "" {
buildsFailed.WithValues(metricsDockerfileEmptyError).Inc()
return nil, errors.New("No image was generated. Is your Dockerfile empty?")
}
return &builder.Result{ImageID: dispatchState.imageID, FromImage: dispatchState.baseImage}, nil
}
func emitImageID(aux *streamformatter.AuxFormatter, state *dispatchState) error {
if aux == nil || state.imageID == "" {
return nil
}
return aux.Emit(types.BuildResult{ID: state.imageID})
}
func (b *Builder) dispatchDockerfileWithCancellation(dockerfile *parser.Result, source builder.Source) (*dispatchState, error) {
shlex := NewShellLex(dockerfile.EscapeToken)
state := newDispatchState()
total := len(dockerfile.AST.Children)
var err error
for i, n := range dockerfile.AST.Children {
select {
case <-b.clientCtx.Done():
logrus.Debug("Builder: build cancelled!")
fmt.Fprint(b.Stdout, "Build cancelled")
buildsFailed.WithValues(metricsBuildCanceled).Inc()
return nil, errors.New("Build cancelled")
default:
// Not cancelled yet, keep going...
}
// If this is a FROM and we have a previous image then
// emit an aux message for that image since it is the
// end of the previous stage
if n.Value == command.From {
if err := emitImageID(b.Aux, state); err != nil {
return nil, err
}
}
if n.Value == command.From && state.isCurrentStage(b.options.Target) {
break
}
opts := dispatchOptions{
state: state,
stepMsg: formatStep(i, total),
node: n,
shlex: shlex,
source: source,
}
if state, err = b.dispatch(opts); err != nil {
if b.options.ForceRemove {
b.containerManager.RemoveAll(b.Stdout)
}
return nil, err
}
fmt.Fprintf(b.Stdout, " ---> %s\n", stringid.TruncateID(state.imageID))
if b.options.Remove {
b.containerManager.RemoveAll(b.Stdout)
}
}
// Emit a final aux message for the final image
if err := emitImageID(b.Aux, state); err != nil {
return nil, err
}
return state, nil
}
func addNodesForLabelOption(dockerfile *parser.Node, labels map[string]string) {
if len(labels) == 0 {
return
}
node := parser.NodeFromLabels(labels)
dockerfile.Children = append(dockerfile.Children, node)
}
// BuildFromConfig builds directly from `changes`, treating it as if it were the contents of a Dockerfile
// It will:
// - Call parse.Parse() to get an AST root for the concatenated Dockerfile entries.
// - Do build by calling builder.dispatch() to call all entries' handling routines
//
// BuildFromConfig is used by the /commit endpoint, with the changes
// coming from the query parameter of the same name.
//
// TODO: Remove?
func BuildFromConfig(config *container.Config, changes []string) (*container.Config, error) {
if len(changes) == 0 {
return config, nil
}
b := newBuilder(context.Background(), builderOptions{
Options: &types.ImageBuildOptions{NoCache: true},
})
dockerfile, err := parser.Parse(bytes.NewBufferString(strings.Join(changes, "\n")))
if err != nil {
return nil, err
}
// ensure that the commands are valid
for _, n := range dockerfile.AST.Children {
if !validCommitCommands[n.Value] {
return nil, fmt.Errorf("%s is not a valid change command", n.Value)
}
}
b.Stdout = ioutil.Discard
b.Stderr = ioutil.Discard
b.disableCommit = true
if err := checkDispatchDockerfile(dockerfile.AST); err != nil {
return nil, err
}
dispatchState := newDispatchState()
dispatchState.runConfig = config
return dispatchFromDockerfile(b, dockerfile, dispatchState, nil)
}
func checkDispatchDockerfile(dockerfile *parser.Node) error {
for _, n := range dockerfile.Children {
if err := checkDispatch(n); err != nil {
return errors.Wrapf(err, "Dockerfile parse error line %d", n.StartLine)
}
}
return nil
}
func dispatchFromDockerfile(b *Builder, result *parser.Result, dispatchState *dispatchState, source builder.Source) (*container.Config, error) {
shlex := NewShellLex(result.EscapeToken)
ast := result.AST
total := len(ast.Children)
for i, n := range ast.Children {
opts := dispatchOptions{
state: dispatchState,
stepMsg: formatStep(i, total),
node: n,
shlex: shlex,
source: source,
}
if _, err := b.dispatch(opts); err != nil {
return nil, err
}
}
return dispatchState.runConfig, nil
}