🐛🎨Do not fail a pipeline when the clusters-keeper or the computational backend in general is not reachable for short time 🚨

services/director-v2/src/simcore_service_director_v2/core/settings.py

@@ -6,6 +6,7 @@
 from functools import cached_property
 from typing import Annotated, cast
 
+from common_library.basic_types import DEFAULT_FACTORY
 from common_library.pydantic_validators import validate_numeric_string_as_timedelta
 from fastapi import FastAPI
 from models_library.basic_types import LogLevel, PortInt
@@ -50,38 +51,53 @@
 
 
 class ComputationalBackendSettings(BaseCustomSettings):
-    COMPUTATIONAL_BACKEND_ENABLED: bool = Field(
-        default=True,
-    )
-    COMPUTATIONAL_BACKEND_SCHEDULING_CONCURRENCY: PositiveInt = Field(
-        default=50,
-        description="defines how many pipelines the application can schedule concurrently",
-    )
-    COMPUTATIONAL_BACKEND_DASK_CLIENT_ENABLED: bool = Field(
-        default=True,
-    )
-    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_URL: AnyUrl = Field(
-        ...,
-        description="This is the cluster that will be used by default"
-        " when submitting computational services (typically "
-        "tcp://dask-scheduler:8786, tls://dask-scheduler:8786 for the internal cluster",
-    )
-    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_AUTH: ClusterAuthentication = Field(
-        default=...,
-        description="this is the cluster authentication that will be used by default",
-    )
-    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_FILE_LINK_TYPE: FileLinkType = Field(
-        FileLinkType.S3,
-        description=f"Default file link type to use with the internal cluster '{list(FileLinkType)}'",
-    )
-    COMPUTATIONAL_BACKEND_DEFAULT_FILE_LINK_TYPE: FileLinkType = Field(
-        FileLinkType.PRESIGNED,
-        description=f"Default file link type to use with computational backend '{list(FileLinkType)}'",
-    )
-    COMPUTATIONAL_BACKEND_ON_DEMAND_CLUSTERS_FILE_LINK_TYPE: FileLinkType = Field(
-        FileLinkType.PRESIGNED,
-        description=f"Default file link type to use with computational backend on-demand clusters '{list(FileLinkType)}'",
-    )
+    COMPUTATIONAL_BACKEND_ENABLED: bool = True
+    COMPUTATIONAL_BACKEND_SCHEDULING_CONCURRENCY: Annotated[
+        PositiveInt,
+        Field(
+            description="defines how many pipelines the application can schedule concurrently"
+        ),
+    ] = 50
+    COMPUTATIONAL_BACKEND_DASK_CLIENT_ENABLED: bool = True
+    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_URL: Annotated[
+        AnyUrl,
+        Field(
+            description="This is the cluster that will be used by default"
+            " when submitting computational services (typically "
+            "tcp://dask-scheduler:8786, tls://dask-scheduler:8786 for the internal cluster",
+        ),
+    ]
+    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_AUTH: Annotated[
+        ClusterAuthentication,
+        Field(
+            description="this is the cluster authentication that will be used by default"
+        ),
+    ]
+    COMPUTATIONAL_BACKEND_DEFAULT_CLUSTER_FILE_LINK_TYPE: Annotated[
+        FileLinkType,
+        Field(
+            description=f"Default file link type to use with the internal cluster '{list(FileLinkType)}'"
+        ),
+    ] = FileLinkType.S3
+    COMPUTATIONAL_BACKEND_DEFAULT_FILE_LINK_TYPE: Annotated[
+        FileLinkType,
+        Field(
+            description=f"Default file link type to use with computational backend '{list(FileLinkType)}'"
+        ),
+    ] = FileLinkType.PRESIGNED
+    COMPUTATIONAL_BACKEND_ON_DEMAND_CLUSTERS_FILE_LINK_TYPE: Annotated[
+        FileLinkType,
+        Field(
+            description=f"Default file link type to use with computational backend on-demand clusters '{list(FileLinkType)}'"
+        ),
+    ] = FileLinkType.PRESIGNED
+    COMPUTATIONAL_BACKEND_MAX_WAITING_FOR_CLUSTER_TIMEOUT: Annotated[
+        datetime.timedelta,
+        Field(
+            description="maximum time a pipeline can wait for a cluster to start"
+            "(default to seconds, or see https://pydantic-docs.helpmanual.io/usage/types/#datetime-types for string formatting)."
+        ),
+    ] = datetime.timedelta(minutes=10)
 
     @cached_property
     def default_cluster(self) -> BaseCluster:
@@ -111,91 +127,107 @@ class AppSettings(BaseApplicationSettings, MixinLoggingSettings):
         ),
     ] = LogLevel.INFO
 
-    DIRECTOR_V2_LOG_FORMAT_LOCAL_DEV_ENABLED: bool = Field(
-        default=False,
-        validation_alias=AliasChoices(
-            "DIRECTOR_V2_LOG_FORMAT_LOCAL_DEV_ENABLED",
-            "LOG_FORMAT_LOCAL_DEV_ENABLED",
+    DIRECTOR_V2_LOG_FORMAT_LOCAL_DEV_ENABLED: Annotated[
+        bool,
+        Field(
+            validation_alias=AliasChoices(
+                "DIRECTOR_V2_LOG_FORMAT_LOCAL_DEV_ENABLED",
+                "LOG_FORMAT_LOCAL_DEV_ENABLED",
+            ),
+            description="Enables local development log format. WARNING: make sure it is disabled if you want to have structured logs!",
         ),
-        description="Enables local development log format. WARNING: make sure it is disabled if you want to have structured logs!",
-    )
-    DIRECTOR_V2_LOG_FILTER_MAPPING: dict[LoggerName, list[MessageSubstring]] = Field(
-        default_factory=dict,
-        validation_alias=AliasChoices(
-            "DIRECTOR_V2_LOG_FILTER_MAPPING", "LOG_FILTER_MAPPING"
+    ] = False
+    DIRECTOR_V2_LOG_FILTER_MAPPING: Annotated[
+        dict[LoggerName, list[MessageSubstring]],
+        Field(
+            default_factory=dict,
+            validation_alias=AliasChoices(
+                "DIRECTOR_V2_LOG_FILTER_MAPPING", "LOG_FILTER_MAPPING"
+            ),
+            description="is a dictionary that maps specific loggers (such as 'uvicorn.access' or 'gunicorn.access') to a list of log message patterns that should be filtered out.",
         ),
-        description="is a dictionary that maps specific loggers (such as 'uvicorn.access' or 'gunicorn.access') to a list of log message patterns that should be filtered out.",
-    )
+    ] = DEFAULT_FACTORY
     DIRECTOR_V2_DEV_FEATURES_ENABLED: bool = False
 
-    DIRECTOR_V2_DEV_FEATURE_R_CLONE_MOUNTS_ENABLED: bool = Field(
-        default=False,
-        description=(
-            "Under development feature. If enabled state "
-            "is saved using rclone docker volumes."
+    DIRECTOR_V2_DEV_FEATURE_R_CLONE_MOUNTS_ENABLED: Annotated[
+        bool,
+        Field(
+            description=(
+                "Under development feature. If enabled state "
+                "is saved using rclone docker volumes."
+            )
         ),
-    )
+    ] = False
 
     # for passing self-signed certificate to spawned services
-    DIRECTOR_V2_SELF_SIGNED_SSL_SECRET_ID: str = Field(
-        default="",
-        description="ID of the docker secret containing the self-signed certificate",
-    )
-    DIRECTOR_V2_SELF_SIGNED_SSL_SECRET_NAME: str = Field(
-        default="",
-        description="Name of the docker secret containing the self-signed certificate",
-    )
-    DIRECTOR_V2_SELF_SIGNED_SSL_FILENAME: str = Field(
-        default="",
-        description="Filepath to self-signed osparc.crt file *as mounted inside the container*, empty strings disables it",
-    )
+    DIRECTOR_V2_SELF_SIGNED_SSL_SECRET_ID: Annotated[
+        str,
+        Field(
+            description="ID of the docker secret containing the self-signed certificate"
+        ),
+    ] = ""
+    DIRECTOR_V2_SELF_SIGNED_SSL_SECRET_NAME: Annotated[
+        str,
+        Field(
+            description="Name of the docker secret containing the self-signed certificate"
+        ),
+    ] = ""
+    DIRECTOR_V2_SELF_SIGNED_SSL_FILENAME: Annotated[
+        str,
+        Field(
+            description="Filepath to self-signed osparc.crt file *as mounted inside the container*, empty strings disables it"
+        ),
+    ] = ""
     DIRECTOR_V2_PROMETHEUS_INSTRUMENTATION_ENABLED: bool = True
     DIRECTOR_V2_PROFILING: bool = False
 
-    DIRECTOR_V2_REMOTE_DEBUGGING_PORT: PortInt | None = Field(default=None)
+    DIRECTOR_V2_REMOTE_DEBUGGING_PORT: PortInt | None = None
 
     # extras
-    SWARM_STACK_NAME: str = Field(default="undefined-please-check")
-    SERVICE_TRACKING_HEARTBEAT: datetime.timedelta = Field(
-        default=DEFAULT_RESOURCE_USAGE_HEARTBEAT_INTERVAL,
-        description="Service scheduler heartbeat (everytime a heartbeat is sent into RabbitMQ)"
-        " (default to seconds, or see https://pydantic-docs.helpmanual.io/usage/types/#datetime-types for string formating)",
-    )
+    SWARM_STACK_NAME: str = "undefined-please-check"
+    SERVICE_TRACKING_HEARTBEAT: Annotated[
+        datetime.timedelta,
+        Field(
+            description="Service scheduler heartbeat (everytime a heartbeat is sent into RabbitMQ)"
+            " (default to seconds, or see https://pydantic-docs.helpmanual.io/usage/types/#datetime-types for string formating)"
+        ),
+    ] = DEFAULT_RESOURCE_USAGE_HEARTBEAT_INTERVAL
 
-    SIMCORE_SERVICES_NETWORK_NAME: str | None = Field(
-        default=None,
-        description="used to find the right network name",
-    )
-    SIMCORE_SERVICES_PREFIX: str | None = Field(
-        "simcore/services",
-        description="useful when developing with an alternative registry namespace",
-    )
+    SIMCORE_SERVICES_NETWORK_NAME: Annotated[
+        str | None, Field(description="used to find the right network name")
+    ] = None
+    SIMCORE_SERVICES_PREFIX: Annotated[
+        str | None,
+        Field(
+            description="useful when developing with an alternative registry namespace"
+        ),
+    ] = "simcore/services"
 
-    DIRECTOR_V2_NODE_PORTS_400_REQUEST_TIMEOUT_ATTEMPTS: NonNegativeInt = Field(
-        default=NODE_PORTS_400_REQUEST_TIMEOUT_ATTEMPTS_DEFAULT_VALUE,
-        description="forwarded to sidecars which use nodeports",
-    )
+    DIRECTOR_V2_NODE_PORTS_400_REQUEST_TIMEOUT_ATTEMPTS: Annotated[
+        NonNegativeInt, Field(description="forwarded to sidecars which use nodeports")
+    ] = NODE_PORTS_400_REQUEST_TIMEOUT_ATTEMPTS_DEFAULT_VALUE
 
     # debug settings
-    CLIENT_REQUEST: ClientRequestSettings = Field(
-        json_schema_extra={"auto_default_from_env": True}
-    )
+    CLIENT_REQUEST: Annotated[
+        ClientRequestSettings, Field(json_schema_extra={"auto_default_from_env": True})
+    ] = DEFAULT_FACTORY
 
     # App modules settings ---------------------
     DIRECTOR_V2_STORAGE: Annotated[
         StorageSettings, Field(json_schema_extra={"auto_default_from_env": True})
     ]
-    DIRECTOR_V2_NODE_PORTS_STORAGE_AUTH: StorageAuthSettings | None = Field(
-        json_schema_extra={"auto_default_from_env": True}
-    )
+    DIRECTOR_V2_NODE_PORTS_STORAGE_AUTH: Annotated[
+        StorageAuthSettings | None,
+        Field(json_schema_extra={"auto_default_from_env": True}),
+    ] = None
 
     DIRECTOR_V2_CATALOG: Annotated[
         CatalogSettings | None, Field(json_schema_extra={"auto_default_from_env": True})
     ]
 
-    DIRECTOR_V0: DirectorV0Settings = Field(
-        json_schema_extra={"auto_default_from_env": True}
-    )
+    DIRECTOR_V0: Annotated[
+        DirectorV0Settings, Field(json_schema_extra={"auto_default_from_env": True})
+    ] = DEFAULT_FACTORY
 
     DYNAMIC_SERVICES: Annotated[
         DynamicServicesSettings,
@@ -206,35 +238,47 @@ class AppSettings(BaseApplicationSettings, MixinLoggingSettings):
         PostgresSettings, Field(json_schema_extra={"auto_default_from_env": True})
     ]
 
-    REDIS: RedisSettings = Field(json_schema_extra={"auto_default_from_env": True})
+    REDIS: Annotated[
+        RedisSettings, Field(json_schema_extra={"auto_default_from_env": True})
+    ] = DEFAULT_FACTORY
 
-    DIRECTOR_V2_RABBITMQ: RabbitSettings = Field(
-        json_schema_extra={"auto_default_from_env": True}
-    )
+    DIRECTOR_V2_RABBITMQ: Annotated[
+        RabbitSettings, Field(json_schema_extra={"auto_default_from_env": True})
+    ] = DEFAULT_FACTORY
 
-    TRAEFIK_SIMCORE_ZONE: str = Field("internal_simcore_stack")
+    TRAEFIK_SIMCORE_ZONE: str = "internal_simcore_stack"
 
-    DIRECTOR_V2_COMPUTATIONAL_BACKEND: ComputationalBackendSettings = Field(
-        json_schema_extra={"auto_default_from_env": True}
-    )
+    DIRECTOR_V2_COMPUTATIONAL_BACKEND: Annotated[
+        ComputationalBackendSettings,
+        Field(json_schema_extra={"auto_default_from_env": True}),
+    ] = DEFAULT_FACTORY
 
-    DIRECTOR_V2_DOCKER_REGISTRY: RegistrySettings = Field(
-        json_schema_extra={"auto_default_from_env": True},
-        description="settings for the private registry deployed with the platform",
-    )
-    DIRECTOR_V2_DOCKER_HUB_REGISTRY: RegistrySettings | None = Field(
-        default=None, description="public DockerHub registry settings"
-    )
+    DIRECTOR_V2_DOCKER_REGISTRY: Annotated[
+        RegistrySettings,
+        Field(
+            json_schema_extra={"auto_default_from_env": True},
+            description="settings for the private registry deployed with the platform",
+        ),
+    ] = DEFAULT_FACTORY
+    DIRECTOR_V2_DOCKER_HUB_REGISTRY: Annotated[
+        RegistrySettings | None, Field(description="public DockerHub registry settings")
+    ] = None
 
-    DIRECTOR_V2_RESOURCE_USAGE_TRACKER: ResourceUsageTrackerSettings = Field(
-        json_schema_extra={"auto_default_from_env": True},
-        description="resource usage tracker service client's plugin",
-    )
+    DIRECTOR_V2_RESOURCE_USAGE_TRACKER: Annotated[
+        ResourceUsageTrackerSettings,
+        Field(
+            json_schema_extra={"auto_default_from_env": True},
+            description="resource usage tracker service client's plugin",
+        ),
+    ] = DEFAULT_FACTORY
 
-    DIRECTOR_V2_TRACING: TracingSettings | None = Field(
-        json_schema_extra={"auto_default_from_env": True},
-        description="settings for opentelemetry tracing",
-    )
+    DIRECTOR_V2_TRACING: Annotated[
+        TracingSettings | None,
+        Field(
+            json_schema_extra={"auto_default_from_env": True},
+            description="settings for opentelemetry tracing",
+        ),
+    ] = None
 
     @field_validator("LOG_LEVEL", mode="before")
     @classmethod

services/director-v2/src/simcore_service_director_v2/models/comp_tasks.py

@@ -276,3 +276,15 @@ class ComputationTaskForRpcDBGet(BaseModel):
     image: dict[str, Any]
     started_at: dt.datetime | None
     ended_at: dt.datetime | None
+
+    @field_validator("state", mode="before")
+    @classmethod
+    def _convert_from_state_type_enum_if_needed(cls, v):
+        if isinstance(v, str):
+            # try to convert to a StateType, if it fails the validations will continue
+            # and pydantic will try to convert it to a RunninState later on
+            with suppress(ValueError):
+                v = StateType(v)
+        if isinstance(v, StateType):
+            return RunningState(DB_TO_RUNNING_STATE[v])
+        return v