Skypilot Kubernetes

zenml.integrations.skypilot_kubernetes

Initialization of the Skypilot Kubernetes integration for ZenML.

The Skypilot integration sub-module powers an alternative to the local orchestrator for a remote orchestration of ZenML pipelines on VMs.

Attributes

SKYPILOT_KUBERNETES = 'skypilot_kubernetes' module-attribute

SKYPILOT_KUBERNETES_ORCHESTRATOR_FLAVOR = 'vm_kubernetes' module-attribute

Classes

Flavor

Class for ZenML Flavors.

Attributes

config_class: Type[StackComponentConfig] abstractmethod property

Returns StackComponentConfig config class.

Returns:

Type	Description
Type[StackComponentConfig]	The config class.

config_schema: Dict[str, Any] property

The config schema for a flavor.

Returns:

Type	Description
Dict[str, Any]	The config schema.

docs_url: Optional[str] property

A url to point at docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor docs url.

implementation_class: Type[StackComponent] abstractmethod property

Implementation class for this flavor.

Returns:

Type	Description
Type[StackComponent]	The implementation class for this flavor.

logo_url: Optional[str] property

A url to represent the flavor in the dashboard.

Returns:

Type	Description
Optional[str]	The flavor logo.

name: str abstractmethod property

The flavor name.

Returns:

Type	Description
str	The flavor name.

sdk_docs_url: Optional[str] property

A url to point at SDK docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor SDK docs url.

service_connector_requirements: Optional[ServiceConnectorRequirements] property

Service connector resource requirements for service connectors.

Specifies resource requirements that are used to filter the available service connector types that are compatible with this flavor.

Returns:

Type	Description
Optional[ServiceConnectorRequirements]	Requirements for compatible service connectors, if a service
Optional[ServiceConnectorRequirements]	connector is required for this flavor.

type: StackComponentType abstractmethod property

The stack component type.

Returns:

Type	Description
StackComponentType	The stack component type.

Functions

from_model(flavor_model: FlavorResponse) -> Flavor classmethod

Loads a flavor from a model.

Parameters:

Name	Type	Description	Default
flavor_model	FlavorResponse	The model to load from.	required

Raises:

Type	Description
CustomFlavorImportError	If the custom flavor can't be imported.
ImportError	If the flavor can't be imported.

Returns:

Type	Description
Flavor	The loaded flavor.

Source code in src/zenml/stack/flavor.py

@classmethod
def from_model(cls, flavor_model: FlavorResponse) -> "Flavor":
    """Loads a flavor from a model.

    Args:
        flavor_model: The model to load from.

    Raises:
        CustomFlavorImportError: If the custom flavor can't be imported.
        ImportError: If the flavor can't be imported.

    Returns:
        The loaded flavor.
    """
    try:
        flavor = source_utils.load(flavor_model.source)()
    except (ModuleNotFoundError, ImportError, NotImplementedError) as err:
        if flavor_model.is_custom:
            flavor_module, _ = flavor_model.source.rsplit(".", maxsplit=1)
            expected_file_path = os.path.join(
                source_utils.get_source_root(),
                flavor_module.replace(".", os.path.sep),
            )
            raise CustomFlavorImportError(
                f"Couldn't import custom flavor {flavor_model.name}: "
                f"{err}. Make sure the custom flavor class "
                f"`{flavor_model.source}` is importable. If it is part of "
                "a library, make sure it is installed. If "
                "it is a local code file, make sure it exists at "
                f"`{expected_file_path}.py`."
            )
        else:
            raise ImportError(
                f"Couldn't import flavor {flavor_model.name}: {err}"
            )
    return cast(Flavor, flavor)

generate_default_docs_url() -> str

Generate the doc urls for all inbuilt and integration flavors.

Note that this method is not going to be useful for custom flavors, which do not have any docs in the main zenml docs.

Returns:

Type	Description
str	The complete url to the zenml documentation

Source code in src/zenml/stack/flavor.py

def generate_default_docs_url(self) -> str:
    """Generate the doc urls for all inbuilt and integration flavors.

    Note that this method is not going to be useful for custom flavors,
    which do not have any docs in the main zenml docs.

    Returns:
        The complete url to the zenml documentation
    """
    from zenml import __version__

    component_type = self.type.plural.replace("_", "-")
    name = self.name.replace("_", "-")

    try:
        is_latest = is_latest_zenml_version()
    except RuntimeError:
        # We assume in error cases that we are on the latest version
        is_latest = True

    if is_latest:
        base = "https://docs.zenml.io"
    else:
        base = f"https://zenml-io.gitbook.io/zenml-legacy-documentation/v/{__version__}"
    return f"{base}/stack-components/{component_type}/{name}"

generate_default_sdk_docs_url() -> str

Generate SDK docs url for a flavor.

Returns:

Type	Description
str	The complete url to the zenml SDK docs

Source code in src/zenml/stack/flavor.py

def generate_default_sdk_docs_url(self) -> str:
    """Generate SDK docs url for a flavor.

    Returns:
        The complete url to the zenml SDK docs
    """
    from zenml import __version__

    base = f"https://sdkdocs.zenml.io/{__version__}"

    component_type = self.type.plural

    if "zenml.integrations" in self.__module__:
        # Get integration name out of module path which will look something
        #  like this "zenml.integrations.<integration>....
        integration = self.__module__.split(
            "zenml.integrations.", maxsplit=1
        )[1].split(".")[0]

        return (
            f"{base}/integration_code_docs"
            f"/integrations-{integration}/#{self.__module__}"
        )

    else:
        return (
            f"{base}/core_code_docs/core-{component_type}/"
            f"#{self.__module__}"
        )

to_model(integration: Optional[str] = None, is_custom: bool = True) -> FlavorRequest

Converts a flavor to a model.

Parameters:

Name	Type	Description	Default
integration	Optional[str]	The integration to use for the model.	None
is_custom	bool	Whether the flavor is a custom flavor. Custom flavors are then scoped by user and workspace	True

Returns:

Type	Description
FlavorRequest	The model.

Source code in src/zenml/stack/flavor.py

def to_model(
    self,
    integration: Optional[str] = None,
    is_custom: bool = True,
) -> FlavorRequest:
    """Converts a flavor to a model.

    Args:
        integration: The integration to use for the model.
        is_custom: Whether the flavor is a custom flavor. Custom flavors
            are then scoped by user and workspace

    Returns:
        The model.
    """
    connector_requirements = self.service_connector_requirements
    connector_type = (
        connector_requirements.connector_type
        if connector_requirements
        else None
    )
    resource_type = (
        connector_requirements.resource_type
        if connector_requirements
        else None
    )
    resource_id_attr = (
        connector_requirements.resource_id_attr
        if connector_requirements
        else None
    )
    user = None
    workspace = None
    if is_custom:
        user = Client().active_user.id
        workspace = Client().active_workspace.id

    model_class = FlavorRequest if is_custom else InternalFlavorRequest
    model = model_class(
        user=user,
        workspace=workspace,
        name=self.name,
        type=self.type,
        source=source_utils.resolve(self.__class__).import_path,
        config_schema=self.config_schema,
        connector_type=connector_type,
        connector_resource_type=resource_type,
        connector_resource_id_attr=resource_id_attr,
        integration=integration,
        logo_url=self.logo_url,
        docs_url=self.docs_url,
        sdk_docs_url=self.sdk_docs_url,
        is_custom=is_custom,
    )
    return model

Integration

Base class for integration in ZenML.

Functions

activate() -> None classmethod

Abstract method to activate the integration.

Source code in src/zenml/integrations/integration.py

@classmethod
def activate(cls) -> None:
    """Abstract method to activate the integration."""

check_installation() -> bool classmethod

Method to check whether the required packages are installed.

Returns:

Type	Description
bool	True if all required packages are installed, False otherwise.

Source code in src/zenml/integrations/integration.py

@classmethod
def check_installation(cls) -> bool:
    """Method to check whether the required packages are installed.

    Returns:
        True if all required packages are installed, False otherwise.
    """
    for r in cls.get_requirements():
        try:
            # First check if the base package is installed
            dist = pkg_resources.get_distribution(r)

            # Next, check if the dependencies (including extras) are
            # installed
            deps: List[Requirement] = []

            _, extras = parse_requirement(r)
            if extras:
                extra_list = extras[1:-1].split(",")
                for extra in extra_list:
                    try:
                        requirements = dist.requires(extras=[extra])  # type: ignore[arg-type]
                    except pkg_resources.UnknownExtra as e:
                        logger.debug(f"Unknown extra: {str(e)}")
                        return False
                    deps.extend(requirements)
            else:
                deps = dist.requires()

            for ri in deps:
                try:
                    # Remove the "extra == ..." part from the requirement string
                    cleaned_req = re.sub(
                        r"; extra == \"\w+\"", "", str(ri)
                    )
                    pkg_resources.get_distribution(cleaned_req)
                except pkg_resources.DistributionNotFound as e:
                    logger.debug(
                        f"Unable to find required dependency "
                        f"'{e.req}' for requirement '{r}' "
                        f"necessary for integration '{cls.NAME}'."
                    )
                    return False
                except pkg_resources.VersionConflict as e:
                    logger.debug(
                        f"Package version '{e.dist}' does not match "
                        f"version '{e.req}' required by '{r}' "
                        f"necessary for integration '{cls.NAME}'."
                    )
                    return False

        except pkg_resources.DistributionNotFound as e:
            logger.debug(
                f"Unable to find required package '{e.req}' for "
                f"integration {cls.NAME}."
            )
            return False
        except pkg_resources.VersionConflict as e:
            logger.debug(
                f"Package version '{e.dist}' does not match version "
                f"'{e.req}' necessary for integration {cls.NAME}."
            )
            return False

    logger.debug(
        f"Integration {cls.NAME} is installed correctly with "
        f"requirements {cls.get_requirements()}."
    )
    return True

flavors() -> List[Type[Flavor]] classmethod

Abstract method to declare new stack component flavors.

Returns:

Type	Description
List[Type[Flavor]]	A list of new stack component flavors.

Source code in src/zenml/integrations/integration.py

@classmethod
def flavors(cls) -> List[Type[Flavor]]:
    """Abstract method to declare new stack component flavors.

    Returns:
        A list of new stack component flavors.
    """
    return []

get_requirements(target_os: Optional[str] = None) -> List[str] classmethod

Method to get the requirements for the integration.

Parameters:

Name	Type	Description	Default
target_os	Optional[str]	The target operating system to get the requirements for.	None

Returns:

Type	Description
List[str]	A list of requirements.

Source code in src/zenml/integrations/integration.py

@classmethod
def get_requirements(cls, target_os: Optional[str] = None) -> List[str]:
    """Method to get the requirements for the integration.

    Args:
        target_os: The target operating system to get the requirements for.

    Returns:
        A list of requirements.
    """
    return cls.REQUIREMENTS

get_uninstall_requirements(target_os: Optional[str] = None) -> List[str] classmethod

Method to get the uninstall requirements for the integration.

Parameters:

Name	Type	Description	Default
target_os	Optional[str]	The target operating system to get the requirements for.	None

Returns:

Type	Description
List[str]	A list of requirements.

Source code in src/zenml/integrations/integration.py

@classmethod
def get_uninstall_requirements(
    cls, target_os: Optional[str] = None
) -> List[str]:
    """Method to get the uninstall requirements for the integration.

    Args:
        target_os: The target operating system to get the requirements for.

    Returns:
        A list of requirements.
    """
    ret = []
    for each in cls.get_requirements(target_os=target_os):
        is_ignored = False
        for ignored in cls.REQUIREMENTS_IGNORED_ON_UNINSTALL:
            if each.startswith(ignored):
                is_ignored = True
                break
        if not is_ignored:
            ret.append(each)
    return ret

plugin_flavors() -> List[Type[BasePluginFlavor]] classmethod

Abstract method to declare new plugin flavors.

Returns:

Type	Description
List[Type[BasePluginFlavor]]	A list of new plugin flavors.

Source code in src/zenml/integrations/integration.py

@classmethod
def plugin_flavors(cls) -> List[Type["BasePluginFlavor"]]:
    """Abstract method to declare new plugin flavors.

    Returns:
        A list of new plugin flavors.
    """
    return []

SkypilotKubernetesIntegration

Bases: Integration

Definition of Skypilot Kubernetes Integration for ZenML.

Functions

flavors() -> List[Type[Flavor]] classmethod

Declare the stack component flavors for the Skypilot Kubernetes integration.

Returns:

Type	Description
List[Type[Flavor]]	List of stack component flavors for this integration.

Source code in src/zenml/integrations/skypilot_kubernetes/__init__.py

@classmethod
def flavors(cls) -> List[Type[Flavor]]:
    """Declare the stack component flavors for the Skypilot Kubernetes integration.

    Returns:
        List of stack component flavors for this integration.
    """
    from zenml.integrations.skypilot_kubernetes.flavors import (
        SkypilotKubernetesOrchestratorFlavor,
    )

    return [SkypilotKubernetesOrchestratorFlavor]

Modules

flavors

Skypilot integration flavor for Skypilot Kubernetes orchestrator.

Classes

SkypilotKubernetesOrchestratorConfig(warn_about_plain_text_secrets: bool = False, **kwargs: Any)

Bases: SkypilotBaseOrchestratorConfig, SkypilotKubernetesOrchestratorSettings

Skypilot orchestrator config.

Source code in src/zenml/stack/stack_component.py

def __init__(
    self, warn_about_plain_text_secrets: bool = False, **kwargs: Any
) -> None:
    """Ensures that secret references don't clash with pydantic validation.

    StackComponents allow the specification of all their string attributes
    using secret references of the form `{{secret_name.key}}`. This however
    is only possible when the stack component does not perform any explicit
    validation of this attribute using pydantic validators. If this were
    the case, the validation would run on the secret reference and would
    fail or in the worst case, modify the secret reference and lead to
    unexpected behavior. This method ensures that no attributes that require
    custom pydantic validation are set as secret references.

    Args:
        warn_about_plain_text_secrets: If true, then warns about using
            plain-text secrets.
        **kwargs: Arguments to initialize this stack component.

    Raises:
        ValueError: If an attribute that requires custom pydantic validation
            is passed as a secret reference, or if the `name` attribute
            was passed as a secret reference.
    """
    for key, value in kwargs.items():
        try:
            field = self.__class__.model_fields[key]
        except KeyError:
            # Value for a private attribute or non-existing field, this
            # will fail during the upcoming pydantic validation
            continue

        if value is None:
            continue

        if not secret_utils.is_secret_reference(value):
            if (
                secret_utils.is_secret_field(field)
                and warn_about_plain_text_secrets
            ):
                logger.warning(
                    "You specified a plain-text value for the sensitive "
                    f"attribute `{key}` for a `{self.__class__.__name__}` "
                    "stack component. This is currently only a warning, "
                    "but future versions of ZenML will require you to pass "
                    "in sensitive information as secrets. Check out the "
                    "documentation on how to configure your stack "
                    "components with secrets here: "
                    "https://docs.zenml.io/getting-started/deploying-zenml/secret-management"
                )
            continue

        if pydantic_utils.has_validators(
            pydantic_class=self.__class__, field_name=key
        ):
            raise ValueError(
                f"Passing the stack component attribute `{key}` as a "
                "secret reference is not allowed as additional validation "
                "is required for this attribute."
            )

    super().__init__(**kwargs)

SkypilotKubernetesOrchestratorFlavor

Bases: BaseOrchestratorFlavor

Flavor for the Skypilot Kubernetes orchestrator.

Attributes

config_class: Type[BaseOrchestratorConfig] property

Config class for the base orchestrator flavor.

Returns:

Type	Description
Type[BaseOrchestratorConfig]	The config class.

docs_url: Optional[str] property

A url to point at docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor docs url.

implementation_class: Type[SkypilotKubernetesOrchestrator] property

Implementation class for this flavor.

Returns:

Type	Description
Type[SkypilotKubernetesOrchestrator]	Implementation class for this flavor.

logo_url: str property

A url to represent the flavor in the dashboard.

Returns:

Type	Description
str	The flavor logo.

name: str property

Name of the orchestrator flavor.

Returns:

Type	Description
str	Name of the orchestrator flavor.

sdk_docs_url: Optional[str] property

A url to point at SDK docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor SDK docs url.

service_connector_requirements: Optional[ServiceConnectorRequirements] property

Service connector resource requirements for service connectors.

Specifies resource requirements that are used to filter the available service connector types that are compatible with this flavor.

Returns:

Type	Description
Optional[ServiceConnectorRequirements]	Requirements for compatible service connectors, if a service
Optional[ServiceConnectorRequirements]	connector is required for this flavor.

SkypilotKubernetesOrchestratorSettings(warn_about_plain_text_secrets: bool = False, **kwargs: Any)

Bases: SkypilotBaseOrchestratorSettings

Skypilot orchestrator settings.

Source code in src/zenml/config/secret_reference_mixin.py

def __init__(
    self, warn_about_plain_text_secrets: bool = False, **kwargs: Any
) -> None:
    """Ensures that secret references are only passed for valid fields.

    This method ensures that secret references are not passed for fields
    that explicitly prevent them or require pydantic validation.

    Args:
        warn_about_plain_text_secrets: If true, then warns about using plain-text secrets.
        **kwargs: Arguments to initialize this object.

    Raises:
        ValueError: If an attribute that requires custom pydantic validation
            or an attribute which explicitly disallows secret references
            is passed as a secret reference.
    """
    for key, value in kwargs.items():
        try:
            field = self.__class__.model_fields[key]
        except KeyError:
            # Value for a private attribute or non-existing field, this
            # will fail during the upcoming pydantic validation
            continue

        if value is None:
            continue

        if not secret_utils.is_secret_reference(value):
            if (
                secret_utils.is_secret_field(field)
                and warn_about_plain_text_secrets
            ):
                logger.warning(
                    "You specified a plain-text value for the sensitive "
                    f"attribute `{key}`. This is currently only a warning, "
                    "but future versions of ZenML will require you to pass "
                    "in sensitive information as secrets. Check out the "
                    "documentation on how to configure values with secrets "
                    "here: https://docs.zenml.io/getting-started/deploying-zenml/secret-management"
                )
            continue

        if secret_utils.is_clear_text_field(field):
            raise ValueError(
                f"Passing the `{key}` attribute as a secret reference is "
                "not allowed."
            )

        requires_validation = has_validators(
            pydantic_class=self.__class__, field_name=key
        )
        if requires_validation:
            raise ValueError(
                f"Passing the attribute `{key}` as a secret reference is "
                "not allowed as additional validation is required for "
                "this attribute."
            )

    super().__init__(**kwargs)

Modules

skypilot_orchestrator_kubernetes_vm_flavor

Skypilot orchestrator Kubernetes flavor.

Classes

SkypilotKubernetesOrchestratorConfig(warn_about_plain_text_secrets: bool = False, **kwargs: Any)

Bases: SkypilotBaseOrchestratorConfig, SkypilotKubernetesOrchestratorSettings

Skypilot orchestrator config.

Source code in src/zenml/stack/stack_component.py

def __init__(
    self, warn_about_plain_text_secrets: bool = False, **kwargs: Any
) -> None:
    """Ensures that secret references don't clash with pydantic validation.

    StackComponents allow the specification of all their string attributes
    using secret references of the form `{{secret_name.key}}`. This however
    is only possible when the stack component does not perform any explicit
    validation of this attribute using pydantic validators. If this were
    the case, the validation would run on the secret reference and would
    fail or in the worst case, modify the secret reference and lead to
    unexpected behavior. This method ensures that no attributes that require
    custom pydantic validation are set as secret references.

    Args:
        warn_about_plain_text_secrets: If true, then warns about using
            plain-text secrets.
        **kwargs: Arguments to initialize this stack component.

    Raises:
        ValueError: If an attribute that requires custom pydantic validation
            is passed as a secret reference, or if the `name` attribute
            was passed as a secret reference.
    """
    for key, value in kwargs.items():
        try:
            field = self.__class__.model_fields[key]
        except KeyError:
            # Value for a private attribute or non-existing field, this
            # will fail during the upcoming pydantic validation
            continue

        if value is None:
            continue

        if not secret_utils.is_secret_reference(value):
            if (
                secret_utils.is_secret_field(field)
                and warn_about_plain_text_secrets
            ):
                logger.warning(
                    "You specified a plain-text value for the sensitive "
                    f"attribute `{key}` for a `{self.__class__.__name__}` "
                    "stack component. This is currently only a warning, "
                    "but future versions of ZenML will require you to pass "
                    "in sensitive information as secrets. Check out the "
                    "documentation on how to configure your stack "
                    "components with secrets here: "
                    "https://docs.zenml.io/getting-started/deploying-zenml/secret-management"
                )
            continue

        if pydantic_utils.has_validators(
            pydantic_class=self.__class__, field_name=key
        ):
            raise ValueError(
                f"Passing the stack component attribute `{key}` as a "
                "secret reference is not allowed as additional validation "
                "is required for this attribute."
            )

    super().__init__(**kwargs)

SkypilotKubernetesOrchestratorFlavor

Bases: BaseOrchestratorFlavor

Flavor for the Skypilot Kubernetes orchestrator.

Attributes

config_class: Type[BaseOrchestratorConfig] property

Config class for the base orchestrator flavor.

Returns:

Type	Description
Type[BaseOrchestratorConfig]	The config class.

docs_url: Optional[str] property

A url to point at docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor docs url.

implementation_class: Type[SkypilotKubernetesOrchestrator] property

Implementation class for this flavor.

Returns:

Type	Description
Type[SkypilotKubernetesOrchestrator]	Implementation class for this flavor.

logo_url: str property

A url to represent the flavor in the dashboard.

Returns:

Type	Description
str	The flavor logo.

name: str property

Name of the orchestrator flavor.

Returns:

Type	Description
str	Name of the orchestrator flavor.

sdk_docs_url: Optional[str] property

A url to point at SDK docs explaining this flavor.

Returns:

Type	Description
Optional[str]	A flavor SDK docs url.

service_connector_requirements: Optional[ServiceConnectorRequirements] property

Service connector resource requirements for service connectors.

Specifies resource requirements that are used to filter the available service connector types that are compatible with this flavor.

Returns:

Type	Description
Optional[ServiceConnectorRequirements]	Requirements for compatible service connectors, if a service
Optional[ServiceConnectorRequirements]	connector is required for this flavor.

SkypilotKubernetesOrchestratorSettings(warn_about_plain_text_secrets: bool = False, **kwargs: Any)

Bases: SkypilotBaseOrchestratorSettings

Skypilot orchestrator settings.

Source code in src/zenml/config/secret_reference_mixin.py

def __init__(
    self, warn_about_plain_text_secrets: bool = False, **kwargs: Any
) -> None:
    """Ensures that secret references are only passed for valid fields.

    This method ensures that secret references are not passed for fields
    that explicitly prevent them or require pydantic validation.

    Args:
        warn_about_plain_text_secrets: If true, then warns about using plain-text secrets.
        **kwargs: Arguments to initialize this object.

    Raises:
        ValueError: If an attribute that requires custom pydantic validation
            or an attribute which explicitly disallows secret references
            is passed as a secret reference.
    """
    for key, value in kwargs.items():
        try:
            field = self.__class__.model_fields[key]
        except KeyError:
            # Value for a private attribute or non-existing field, this
            # will fail during the upcoming pydantic validation
            continue

        if value is None:
            continue

        if not secret_utils.is_secret_reference(value):
            if (
                secret_utils.is_secret_field(field)
                and warn_about_plain_text_secrets
            ):
                logger.warning(
                    "You specified a plain-text value for the sensitive "
                    f"attribute `{key}`. This is currently only a warning, "
                    "but future versions of ZenML will require you to pass "
                    "in sensitive information as secrets. Check out the "
                    "documentation on how to configure values with secrets "
                    "here: https://docs.zenml.io/getting-started/deploying-zenml/secret-management"
                )
            continue

        if secret_utils.is_clear_text_field(field):
            raise ValueError(
                f"Passing the `{key}` attribute as a secret reference is "
                "not allowed."
            )

        requires_validation = has_validators(
            pydantic_class=self.__class__, field_name=key
        )
        if requires_validation:
            raise ValueError(
                f"Passing the attribute `{key}` as a secret reference is "
                "not allowed as additional validation is required for "
                "this attribute."
            )

    super().__init__(**kwargs)

Functions

orchestrators

Initialization of the Skypilot Kubernetes ZenML orchestrator.

Classes

SkypilotBaseOrchestrator(name: str, id: UUID, config: StackComponentConfig, flavor: str, type: StackComponentType, user: Optional[UUID], workspace: UUID, created: datetime, updated: datetime, labels: Optional[Dict[str, Any]] = None, connector_requirements: Optional[ServiceConnectorRequirements] = None, connector: Optional[UUID] = None, connector_resource_id: Optional[str] = None, *args: Any, **kwargs: Any)

Bases: ContainerizedOrchestrator

Base class for Orchestrator responsible for running pipelines remotely in a VM.

This orchestrator does not support running on a schedule.

Source code in src/zenml/stack/stack_component.py

def __init__(
    self,
    name: str,
    id: UUID,
    config: StackComponentConfig,
    flavor: str,
    type: StackComponentType,
    user: Optional[UUID],
    workspace: UUID,
    created: datetime,
    updated: datetime,
    labels: Optional[Dict[str, Any]] = None,
    connector_requirements: Optional[ServiceConnectorRequirements] = None,
    connector: Optional[UUID] = None,
    connector_resource_id: Optional[str] = None,
    *args: Any,
    **kwargs: Any,
):
    """Initializes a StackComponent.

    Args:
        name: The name of the component.
        id: The unique ID of the component.
        config: The config of the component.
        flavor: The flavor of the component.
        type: The type of the component.
        user: The ID of the user who created the component.
        workspace: The ID of the workspace the component belongs to.
        created: The creation time of the component.
        updated: The last update time of the component.
        labels: The labels of the component.
        connector_requirements: The requirements for the connector.
        connector: The ID of a connector linked to the component.
        connector_resource_id: The custom resource ID to access through
            the connector.
        *args: Additional positional arguments.
        **kwargs: Additional keyword arguments.

    Raises:
        ValueError: If a secret reference is passed as name.
    """
    if secret_utils.is_secret_reference(name):
        raise ValueError(
            "Passing the `name` attribute of a stack component as a "
            "secret reference is not allowed."
        )

    self.id = id
    self.name = name
    self._config = config
    self.flavor = flavor
    self.type = type
    self.user = user
    self.workspace = workspace
    self.created = created
    self.updated = updated
    self.labels = labels
    self.connector_requirements = connector_requirements
    self.connector = connector
    self.connector_resource_id = connector_resource_id
    self._connector_instance: Optional[ServiceConnector] = None

Attributes

cloud: sky.clouds.Cloud abstractmethod property

The type of sky cloud to use.

Returns:

Type	Description
Cloud	A sky.clouds.Cloud instance.

config: SkypilotBaseOrchestratorConfig property

Returns the SkypilotBaseOrchestratorConfig config.

Returns:

Type	Description
SkypilotBaseOrchestratorConfig	The configuration.

validator: Optional[StackValidator] property

Validates the stack.

In the remote case, checks that the stack contains a container registry, image builder and only remote components.

Returns:

Type	Description
Optional[StackValidator]	A StackValidator instance.

Functions

get_orchestrator_run_id() -> str

Returns the active orchestrator run id.

Raises:

Type	Description
RuntimeError	If the environment variable specifying the run id is not set.

Returns:

Type	Description
str	The orchestrator run id.

Source code in src/zenml/integrations/skypilot/orchestrators/skypilot_base_vm_orchestrator.py

def get_orchestrator_run_id(self) -> str:
    """Returns the active orchestrator run id.

    Raises:
        RuntimeError: If the environment variable specifying the run id
            is not set.

    Returns:
        The orchestrator run id.
    """
    try:
        return os.environ[ENV_ZENML_SKYPILOT_ORCHESTRATOR_RUN_ID]
    except KeyError:
        raise RuntimeError(
            "Unable to read run id from environment variable "
            f"{ENV_ZENML_SKYPILOT_ORCHESTRATOR_RUN_ID}."
        )

prepare_environment_variable(set: bool = True) -> None abstractmethod

Set up Environment variables that are required for the orchestrator.

Parameters:

Name	Type	Description	Default
set	bool	Whether to set the environment variables or not.	True

Source code in src/zenml/integrations/skypilot/orchestrators/skypilot_base_vm_orchestrator.py

@abstractmethod
def prepare_environment_variable(self, set: bool = True) -> None:
    """Set up Environment variables that are required for the orchestrator.

    Args:
        set: Whether to set the environment variables or not.
    """

prepare_or_run_pipeline(deployment: PipelineDeploymentResponse, stack: Stack, environment: Dict[str, str]) -> Any

Runs each pipeline step in a separate Skypilot container.

Parameters:

Name	Type	Description	Default
deployment	PipelineDeploymentResponse	The pipeline deployment to prepare or run.	required
stack	Stack	The stack the pipeline will run on.	required
environment	Dict[str, str]	Environment variables to set in the orchestration environment.	required

Raises:

Type	Description
Exception	If the pipeline run fails.
RuntimeError	If the code is running in a notebook.

Source code in src/zenml/integrations/skypilot/orchestrators/skypilot_base_vm_orchestrator.py

def prepare_or_run_pipeline(
    self,
    deployment: "PipelineDeploymentResponse",
    stack: "Stack",
    environment: Dict[str, str],
) -> Any:
    """Runs each pipeline step in a separate Skypilot container.

    Args:
        deployment: The pipeline deployment to prepare or run.
        stack: The stack the pipeline will run on.
        environment: Environment variables to set in the orchestration
            environment.

    Raises:
        Exception: If the pipeline run fails.
        RuntimeError: If the code is running in a notebook.
    """
    # First check whether the code is running in a notebook.
    if Environment.in_notebook():
        raise RuntimeError(
            "The Skypilot orchestrator cannot run pipelines in a notebook "
            "environment. The reason is that it is non-trivial to create "
            "a Docker image of a notebook. Please consider refactoring "
            "your notebook cells into separate scripts in a Python module "
            "and run the code outside of a notebook when using this "
            "orchestrator."
        )
    if deployment.schedule:
        logger.warning(
            "Skypilot Orchestrator currently does not support the "
            "use of schedules. The `schedule` will be ignored "
            "and the pipeline will be run immediately."
        )

    # Set up some variables for configuration
    orchestrator_run_id = str(uuid4())
    environment[ENV_ZENML_SKYPILOT_ORCHESTRATOR_RUN_ID] = (
        orchestrator_run_id
    )

    settings = cast(
        SkypilotBaseOrchestratorSettings,
        self.get_settings(deployment),
    )

    pipeline_name = deployment.pipeline_configuration.name
    orchestrator_run_name = get_orchestrator_run_name(pipeline_name)

    assert stack.container_registry

    # Get Docker image for the orchestrator pod
    try:
        image = self.get_image(deployment=deployment)
    except KeyError:
        # If no generic pipeline image exists (which means all steps have
        # custom builds) we use a random step image as all of them include
        # dependencies for the active stack
        pipeline_step_name = next(iter(deployment.step_configurations))
        image = self.get_image(
            deployment=deployment, step_name=pipeline_step_name
        )

    different_settings_found = False

    if not self.config.disable_step_based_settings:
        for _, step in deployment.step_configurations.items():
            step_settings = cast(
                SkypilotBaseOrchestratorSettings,
                self.get_settings(step),
            )
            if step_settings != settings:
                different_settings_found = True
                logger.info(
                    "At least one step has different settings than the "
                    "pipeline. The step with different settings will be "
                    "run in a separate VM.\n"
                    "You can configure the orchestrator to disable this "
                    "behavior by updating the `disable_step_based_settings` "
                    "in your orchestrator configuration "
                    "by running the following command: "
                    "`zenml orchestrator update --disable-step-based-settings=True`"
                )
                break

    # Decide which configuration to use based on whether different settings were found
    if (
        not self.config.disable_step_based_settings
        and different_settings_found
    ):
        # Run each step in a separate VM using SkypilotOrchestratorEntrypointConfiguration
        command = SkypilotOrchestratorEntrypointConfiguration.get_entrypoint_command()
        args = SkypilotOrchestratorEntrypointConfiguration.get_entrypoint_arguments(
            run_name=orchestrator_run_name,
            deployment_id=deployment.id,
        )
    else:
        # Run the entire pipeline in one VM using PipelineEntrypointConfiguration
        command = PipelineEntrypointConfiguration.get_entrypoint_command()
        args = PipelineEntrypointConfiguration.get_entrypoint_arguments(
            deployment_id=deployment.id
        )

    entrypoint_str = " ".join(command)
    arguments_str = " ".join(args)

    task_envs = environment
    docker_environment_str = " ".join(
        f"-e {k}={v}" for k, v in environment.items()
    )
    custom_run_args = " ".join(settings.docker_run_args)
    if custom_run_args:
        custom_run_args += " "

    instance_type = settings.instance_type or self.DEFAULT_INSTANCE_TYPE

    # Set up credentials
    self.setup_credentials()

    # Guaranteed by stack validation
    assert stack is not None and stack.container_registry is not None

    if docker_creds := stack.container_registry.credentials:
        docker_username, docker_password = docker_creds
        setup = (
            f"sudo docker login --username $DOCKER_USERNAME --password "
            f"$DOCKER_PASSWORD {stack.container_registry.config.uri}"
        )
        task_envs["DOCKER_USERNAME"] = docker_username
        task_envs["DOCKER_PASSWORD"] = docker_password
    else:
        setup = None

    # Run the entire pipeline

    # Set the service connector AWS profile ENV variable
    self.prepare_environment_variable(set=True)

    try:
        if isinstance(self.cloud, sky.clouds.Kubernetes):
            run_command = f"${{VIRTUAL_ENV:+$VIRTUAL_ENV/bin/}}{entrypoint_str} {arguments_str}"
            setup = None
            down = False
            idle_minutes_to_autostop = None
        else:
            run_command = f"sudo docker run --rm {custom_run_args}{docker_environment_str} {image} {entrypoint_str} {arguments_str}"
            down = settings.down
            idle_minutes_to_autostop = settings.idle_minutes_to_autostop
        task = sky.Task(
            run=run_command,
            setup=setup,
            envs=task_envs,
        )
        logger.debug(f"Running run: {run_command}")

        task = task.set_resources(
            sky.Resources(
                cloud=self.cloud,
                instance_type=instance_type,
                cpus=settings.cpus,
                memory=settings.memory,
                accelerators=settings.accelerators,
                accelerator_args=settings.accelerator_args,
                use_spot=settings.use_spot,
                job_recovery=settings.job_recovery,
                region=settings.region,
                zone=settings.zone,
                image_id=image
                if isinstance(self.cloud, sky.clouds.Kubernetes)
                else settings.image_id,
                disk_size=settings.disk_size,
                disk_tier=settings.disk_tier,
            )
        )
        # Set the cluster name
        if settings.cluster_name:
            sky.exec(
                task,
                settings.cluster_name,
                down=down,
                stream_logs=settings.stream_logs,
                backend=None,
                detach_run=True,
            )
        else:
            # Find existing cluster
            for i in sky.status(refresh=True):
                if isinstance(
                    i["handle"].launched_resources.cloud, type(self.cloud)
                ):
                    cluster_name = i["handle"].cluster_name
                    logger.info(
                        f"Found existing cluster {cluster_name}. Reusing..."
                    )
            cluster_name = self.sanitize_cluster_name(
                f"{orchestrator_run_name}"
            )
            # Launch the cluster
            sky.launch(
                task,
                cluster_name,
                retry_until_up=settings.retry_until_up,
                idle_minutes_to_autostop=idle_minutes_to_autostop,
                down=down,
                stream_logs=settings.stream_logs,
                detach_setup=True,
            )

    except Exception as e:
        logger.error(f"Pipeline run failed: {e}")
        raise

    finally:
        # Unset the service connector AWS profile ENV variable
        self.prepare_environment_variable(set=False)

sanitize_cluster_name(name: str) -> str

Sanitize the value to be used in a cluster name.

Parameters:

Name	Type	Description	Default
name	str	Arbitrary input cluster name.	required

Returns:

Type	Description
str	Sanitized cluster name.

Source code in src/zenml/integrations/skypilot/orchestrators/skypilot_base_vm_orchestrator.py

def sanitize_cluster_name(self, name: str) -> str:
    """Sanitize the value to be used in a cluster name.

    Args:
        name: Arbitrary input cluster name.

    Returns:
        Sanitized cluster name.
    """
    name = re.sub(
        r"[^a-z0-9-]", "-", name.lower()
    )  # replaces any character that is not a lowercase letter, digit, or hyphen with a hyphen
    name = re.sub(r"^[-]+", "", name)  # trim leading hyphens
    name = re.sub(r"[-]+$", "", name)  # trim trailing hyphens
    return name

setup_credentials() -> None

Set up credentials for the orchestrator.

Source code in src/zenml/integrations/skypilot/orchestrators/skypilot_base_vm_orchestrator.py

def setup_credentials(self) -> None:
    """Set up credentials for the orchestrator."""
    connector = self.get_connector()
    assert connector is not None
    connector.configure_local_client()

SkypilotKubernetesOrchestrator(name: str, id: UUID, config: StackComponentConfig, flavor: str, type: StackComponentType, user: Optional[UUID], workspace: UUID, created: datetime, updated: datetime, labels: Optional[Dict[str, Any]] = None, connector_requirements: Optional[ServiceConnectorRequirements] = None, connector: Optional[UUID] = None, connector_resource_id: Optional[str] = None, *args: Any, **kwargs: Any)

Bases: SkypilotBaseOrchestrator

Orchestrator responsible for running pipelines remotely in a VM on Kubernetes.

This orchestrator does not support running on a schedule.

Source code in src/zenml/stack/stack_component.py

def __init__(
    self,
    name: str,
    id: UUID,
    config: StackComponentConfig,
    flavor: str,
    type: StackComponentType,
    user: Optional[UUID],
    workspace: UUID,
    created: datetime,
    updated: datetime,
    labels: Optional[Dict[str, Any]] = None,
    connector_requirements: Optional[ServiceConnectorRequirements] = None,
    connector: Optional[UUID] = None,
    connector_resource_id: Optional[str] = None,
    *args: Any,
    **kwargs: Any,
):
    """Initializes a StackComponent.

    Args:
        name: The name of the component.
        id: The unique ID of the component.
        config: The config of the component.
        flavor: The flavor of the component.
        type: The type of the component.
        user: The ID of the user who created the component.
        workspace: The ID of the workspace the component belongs to.
        created: The creation time of the component.
        updated: The last update time of the component.
        labels: The labels of the component.
        connector_requirements: The requirements for the connector.
        connector: The ID of a connector linked to the component.
        connector_resource_id: The custom resource ID to access through
            the connector.
        *args: Additional positional arguments.
        **kwargs: Additional keyword arguments.

    Raises:
        ValueError: If a secret reference is passed as name.
    """
    if secret_utils.is_secret_reference(name):
        raise ValueError(
            "Passing the `name` attribute of a stack component as a "
            "secret reference is not allowed."
        )

    self.id = id
    self.name = name
    self._config = config
    self.flavor = flavor
    self.type = type
    self.user = user
    self.workspace = workspace
    self.created = created
    self.updated = updated
    self.labels = labels
    self.connector_requirements = connector_requirements
    self.connector = connector
    self.connector_resource_id = connector_resource_id
    self._connector_instance: Optional[ServiceConnector] = None

Attributes

cloud: sky.clouds.Cloud property

The type of sky cloud to use.

Returns:

Type	Description
Cloud	A sky.clouds.Cloud instance.

config: SkypilotKubernetesOrchestratorConfig property

Returns the SkypilotKubernetesOrchestratorConfig config.

Returns:

Type	Description
SkypilotKubernetesOrchestratorConfig	The configuration.

settings_class: Optional[Type[BaseSettings]] property

Settings class for the Skypilot orchestrator.

Returns:

Type	Description
Optional[Type[BaseSettings]]	The settings class.

Functions

prepare_environment_variable(set: bool = True) -> None

Set up Environment variables that are required for the orchestrator.

Parameters:

Name	Type	Description	Default
set	bool	Whether to set the environment variables or not.	True

Source code in src/zenml/integrations/skypilot_kubernetes/orchestrators/skypilot_kubernetes_vm_orchestrator.py

def prepare_environment_variable(self, set: bool = True) -> None:
    """Set up Environment variables that are required for the orchestrator.

    Args:
        set: Whether to set the environment variables or not.
    """
    pass

Modules

skypilot_kubernetes_vm_orchestrator

Implementation of the a Skypilot based Kubernetes VM orchestrator.

Classes

SkypilotKubernetesOrchestrator(name: str, id: UUID, config: StackComponentConfig, flavor: str, type: StackComponentType, user: Optional[UUID], workspace: UUID, created: datetime, updated: datetime, labels: Optional[Dict[str, Any]] = None, connector_requirements: Optional[ServiceConnectorRequirements] = None, connector: Optional[UUID] = None, connector_resource_id: Optional[str] = None, *args: Any, **kwargs: Any)

Bases: SkypilotBaseOrchestrator

Orchestrator responsible for running pipelines remotely in a VM on Kubernetes.

This orchestrator does not support running on a schedule.

Source code in src/zenml/stack/stack_component.py

def __init__(
    self,
    name: str,
    id: UUID,
    config: StackComponentConfig,
    flavor: str,
    type: StackComponentType,
    user: Optional[UUID],
    workspace: UUID,
    created: datetime,
    updated: datetime,
    labels: Optional[Dict[str, Any]] = None,
    connector_requirements: Optional[ServiceConnectorRequirements] = None,
    connector: Optional[UUID] = None,
    connector_resource_id: Optional[str] = None,
    *args: Any,
    **kwargs: Any,
):
    """Initializes a StackComponent.

    Args:
        name: The name of the component.
        id: The unique ID of the component.
        config: The config of the component.
        flavor: The flavor of the component.
        type: The type of the component.
        user: The ID of the user who created the component.
        workspace: The ID of the workspace the component belongs to.
        created: The creation time of the component.
        updated: The last update time of the component.
        labels: The labels of the component.
        connector_requirements: The requirements for the connector.
        connector: The ID of a connector linked to the component.
        connector_resource_id: The custom resource ID to access through
            the connector.
        *args: Additional positional arguments.
        **kwargs: Additional keyword arguments.

    Raises:
        ValueError: If a secret reference is passed as name.
    """
    if secret_utils.is_secret_reference(name):
        raise ValueError(
            "Passing the `name` attribute of a stack component as a "
            "secret reference is not allowed."
        )

    self.id = id
    self.name = name
    self._config = config
    self.flavor = flavor
    self.type = type
    self.user = user
    self.workspace = workspace
    self.created = created
    self.updated = updated
    self.labels = labels
    self.connector_requirements = connector_requirements
    self.connector = connector
    self.connector_resource_id = connector_resource_id
    self._connector_instance: Optional[ServiceConnector] = None

Attributes

cloud: sky.clouds.Cloud property

The type of sky cloud to use.

Returns:

Type	Description
Cloud	A sky.clouds.Cloud instance.

config: SkypilotKubernetesOrchestratorConfig property

Returns the SkypilotKubernetesOrchestratorConfig config.

Returns:

Type	Description
SkypilotKubernetesOrchestratorConfig	The configuration.

settings_class: Optional[Type[BaseSettings]] property

Settings class for the Skypilot orchestrator.

Returns:

Type	Description
Optional[Type[BaseSettings]]	The settings class.

Functions

prepare_environment_variable(set: bool = True) -> None

Set up Environment variables that are required for the orchestrator.

Parameters:

Name	Type	Description	Default
set	bool	Whether to set the environment variables or not.	True

Source code in src/zenml/integrations/skypilot_kubernetes/orchestrators/skypilot_kubernetes_vm_orchestrator.py

def prepare_environment_variable(self, set: bool = True) -> None:
    """Set up Environment variables that are required for the orchestrator.

    Args:
        set: Whether to set the environment variables or not.
    """
    pass

Functions