Documentation ¶
Overview ¶
Package dockerutil contains helpers for interacting with Docker containers.
Index ¶
- Constants
- Variables
- func CondenseHostName(name string) string
- func CopyFile(src, dst string) (int64, error)
- func DockerSetup(t DockerSetupTestingT) (*client.Client, string)
- func GetDockerUserString() string
- func GetHeighlinerUserString() string
- func GetHostPort(cont types.ContainerJSON, portID string) string
- func GetRootUserString() string
- func NewLocalKeyringFromDockerContainer(ctx context.Context, dc *client.Client, ...) (keyring.Keyring, error)
- func RandLowerCaseLetterString(length int) string
- func SanitizeContainerName(name string) string
- func SetVolumeOwner(ctx context.Context, opts VolumeOwnerOptions) error
- func StartContainer(ctx context.Context, cli *client.Client, id string) error
- type Container
- type ContainerExecResult
- type ContainerOptions
- type DockerSetupTestingT
- type FileRetriever
- type FileWriter
- type Image
- type VolumeOwnerOptions
Constants ¶
const ( // LabelPrefix is the reverse DNS format "namespace" for interchaintest Docker labels. LabelPrefix = "ventures.strangelove.interchaintest." // NodeOwnerLabel indicates the logical node owning a particular object (probably a volume). NodeOwnerLabel = LabelPrefix + "node-owner" )
const CleanupLabel = "ibc-test"
CleanupLabel is a docker label key targeted by DockerSetup when it cleans up docker resources.
"interchaintest" is perhaps a better name. However, for backwards compatability we preserve the original name of "ibc-test" with the hyphen. Otherwise, we run the risk of causing "container already exists" errors because DockerSetup is unable to clean old resources from docker engine.
Variables ¶
var KeepVolumesOnFailure = os.Getenv("IBCTEST_SKIP_FAILURE_CLEANUP") != ""
KeepVolumesOnFailure determines whether volumes associated with a test using DockerSetup are retained or deleted following a test failure.
The value is false by default, but can be initialized to true by setting the environment variable IBCTEST_SKIP_FAILURE_CLEANUP to a non-empty value. Alternatively, importers of the dockerutil package may set the variable to true. Because dockerutil is an internal package, the public API for setting this value is interchaintest.KeepDockerVolumesOnFailure(bool).
Functions ¶
func CondenseHostName ¶
CondenseHostName truncates the middle of the given name if it is 64 characters or longer.
Without this helper, you may see an error like:
API error (500): failed to create shim: OCI runtime create failed: container_linux.go:380: starting container process caused: process_linux.go:545: container init caused: sethostname: invalid argument: unknown
func DockerSetup ¶
func DockerSetup(t DockerSetupTestingT) (*client.Client, string)
DockerSetup returns a new Docker Client and the ID of a configured network, associated with t.
If any part of the setup fails, DockerSetup panics because the test cannot continue.
func GetDockerUserString ¶
func GetDockerUserString() string
func GetHeighlinerUserString ¶
func GetHeighlinerUserString() string
func GetHostPort ¶
func GetHostPort(cont types.ContainerJSON, portID string) string
GetHostPort returns a resource's published port with an address. cont is the type returned by the Docker client's ContainerInspect method.
func GetRootUserString ¶
func GetRootUserString() string
func NewLocalKeyringFromDockerContainer ¶
func NewLocalKeyringFromDockerContainer(ctx context.Context, dc *client.Client, localDirectory, containerKeyringDir, containerId string) (keyring.Keyring, error)
NewLocalKeyringFromDockerContainer copies the contents of the given container directory into a specified local directory. This allows test hosts to sign transactions on behalf of test users.
func RandLowerCaseLetterString ¶
RandLowerCaseLetterString returns a lowercase letter string of given length
func SanitizeContainerName ¶
SanitizeContainerName returns name with any invalid characters replaced with underscores. Subtests will include slashes, and there may be other invalid characters too.
func SetVolumeOwner ¶
func SetVolumeOwner(ctx context.Context, opts VolumeOwnerOptions) error
SetVolumeOwner configures the owner of a volume to match the default user in the supplied image reference.
Types ¶
type Container ¶
Container is a docker container. Use (*Image).Start to create a new container.
func (*Container) Stop ¶
Stop gives the container up to timeout to stop and remove itself from the network.
func (*Container) Wait ¶
func (c *Container) Wait(ctx context.Context, logTail uint64) ContainerExecResult
Wait blocks until the container exits. Calling wait is not suitable for daemons and servers. A non-zero status code returns an error.
Wait implicitly calls Stop. If logTail is non-zero, the stdout and stderr logs will be truncated at the end to that number of lines.
type ContainerExecResult ¶
type ContainerExecResult struct { Err error // Err is nil, unless some error occurs during the container lifecycle. ExitCode int Stdout, Stderr []byte }
ContainerExecResult is a wrapper type that wraps an exit code and associated output from stderr & stdout, along with an error in the case of some error occurring during container execution.
type ContainerOptions ¶
type ContainerOptions struct { // bind mounts: https://docs.docker.com/storage/bind-mounts/ Binds []string // Environment variables Env []string // If blank, defaults to the container's default user. User string // If non-zero, will limit the amount of log lines returned. LogTail uint64 }
ContainerOptions optionally configures starting a Container.
type DockerSetupTestingT ¶
type DockerSetupTestingT interface { Helper() Name() string Failed() bool Cleanup(func()) Logf(format string, args ...any) }
DockerSetupTestingT is a subset of testing.T required for DockerSetup.
type FileRetriever ¶
type FileRetriever struct {
// contains filtered or unexported fields
}
FileRetriever allows retrieving a single file from a Docker volume. In the future it may allow retrieving an entire directory.
func NewFileRetriever ¶
NewFileRetriever returns a new FileRetriever.
func (*FileRetriever) SingleFileContent ¶
func (r *FileRetriever) SingleFileContent(ctx context.Context, volumeName, relPath string) ([]byte, error)
SingleFileContent returns the content of the file named at relPath, inside the volume specified by volumeName.
type FileWriter ¶
type FileWriter struct {
// contains filtered or unexported fields
}
FileWriter allows retrieving a single file from a Docker volume. In the future it may allow retrieving an entire directory.
func NewFileWriter ¶
NewFileWriter returns a new FileWriter.
type Image ¶
type Image struct {
// contains filtered or unexported fields
}
Image is a docker image.
func NewImage ¶
func NewImage(logger *zap.Logger, cli *client.Client, networkID string, testName string, repository, tag string) *Image
NewImage returns a valid Image.
"pool" and "networkID" are likely from DockerSetup. "testName" is from a (*testing.T).Name() and should match the t.Name() from DockerSetup to ensure proper cleanup.
Most arguments (except tag) must be non-zero values or this function panics. If tag is absent, defaults to "latest". Currently, only public docker images are supported.
func (*Image) Run ¶
func (image *Image) Run(ctx context.Context, cmd []string, opts ContainerOptions) ContainerExecResult
Run creates and runs a container invoking "cmd". The container resources are removed after exit.
Run blocks until the command completes. Thus, Run is not suitable for daemons or servers. Use Start instead. A non-zero status code returns an error.