Files
lda-wf/src/wf_api/run_lifecycle.py
T

275 lines
8.4 KiB
Python

from __future__ import annotations
from datetime import UTC, datetime
from uuid import uuid4
from wf_api.saved_subgraphs import SavedSubgraphTree
from wf_artifacts import (
AvailableSource,
CheckpointReason,
DependencyDiagnostic,
DiagnosticSeverity,
PinnedRunEnvironment,
ResumeReadiness,
RunAdmission,
RunCheckpoint,
RunStore,
StoredRunStatus,
WorkflowArtifact,
WorkflowDeployment,
WorkflowRunRecord,
validate_deployment_dependencies,
)
from wf_core import (
PersistedRunState,
RunState,
RunStatus,
dump_run_state,
load_run_state,
load_run_state_with_upgrade,
)
def create_pinned_environment(
*,
deployment: WorkflowDeployment,
artifact: WorkflowArtifact,
tree: SavedSubgraphTree,
) -> PinnedRunEnvironment:
"""Capture exact root, deployment, and child definitions for one run."""
return PinnedRunEnvironment(
deployment=deployment,
root_artifact=artifact,
child_artifacts=list(tree.artifacts_by_ref.values()),
)
def persist_admission(
*,
store: RunStore,
run_id: str,
environment: PinnedRunEnvironment,
resolved_input: dict[str, object],
max_steps: int | None,
scheduled_at: datetime | None = None,
schedule_id: str | None = None,
schedule_revision: int | None = None,
deployment_revision: int | None = None,
) -> RunAdmission:
"""Persist the authoritative admission record before dispatch.
The occurrence is decided here: the caller must treat this persist as the
single authority for the run identity and frozen invocation. A failed
durable admission must never dispatch.
"""
from wf_core.models.json_values import validate_strict_json_value
frozen = validate_strict_json_value(dict(resolved_input))
if not isinstance(frozen, dict):
raise ValueError("resolved workflow input must be a JSON object")
admission = RunAdmission(
id=run_id,
environment=environment,
resolved_input=frozen,
max_steps=max_steps,
scheduled_at=scheduled_at,
schedule_id=schedule_id,
schedule_revision=schedule_revision,
deployment_revision=deployment_revision,
created_at=datetime.now(UTC),
)
store.save_admission(admission)
return admission
def materialize_admitted_view(
*,
store: RunStore,
admission: RunAdmission,
) -> WorkflowRunRecord:
"""Materialize the run admission view using the admission identity.
The view carries no checkpoint, trace, output, or step counts: the outcome
is unknown until dispatch completes and persists a stopped checkpoint.
"""
now = datetime.now(UTC)
try:
existing = store.get_run(admission.id)
except KeyError:
existing = None
if existing is not None:
return existing
record = WorkflowRunRecord(
id=admission.id,
status=StoredRunStatus.ADMITTED,
resume_readiness=ResumeReadiness.NOT_APPLICABLE,
environment=admission.environment,
latest_checkpoint_id=None,
created_at=admission.created_at,
updated_at=now,
)
store.save_run(record)
return record
def recover_admission_view(*, store: RunStore, run_id: str) -> WorkflowRunRecord:
"""Reconcile a missing run view from its admission record.
Recovery never executes work: it only completes the missing view so a
later poll can dispatch the captured invocation exactly once.
"""
try:
return store.get_run(run_id)
except KeyError:
pass
admission = store.get_admission(run_id)
return materialize_admitted_view(store=store, admission=admission)
def persist_stopped_run(
*,
store: RunStore,
environment: PinnedRunEnvironment,
run: RunState,
run_id: str | None = None,
) -> WorkflowRunRecord:
"""Persist one externally visible stopped state and its typed checkpoint."""
if run.status not in {
RunStatus.INTERRUPTED,
RunStatus.COMPLETED,
RunStatus.FAILED,
}:
raise ValueError(
f"cannot persist active workflow run with status {run.status!s}"
)
key = run_id or f"run_{uuid4().hex}"
now = datetime.now(UTC)
sequence = 1
created_at = now
if run_id is not None:
try:
existing = store.get_run(run_id)
except KeyError:
existing = None
if existing is not None:
created_at = existing.created_at
try:
sequence = store.get_latest_checkpoint(run_id).sequence + 1
except KeyError:
sequence = 1
status = StoredRunStatus(run.status.value)
readiness = (
ResumeReadiness.READY
if status is StoredRunStatus.INTERRUPTED
else ResumeReadiness.NOT_APPLICABLE
)
checkpoint_id = f"{key}.{sequence:06d}"
checkpoint = RunCheckpoint(
id=checkpoint_id,
run_id=key,
sequence=sequence,
reason=CheckpointReason(status.value),
state=PersistedRunState.model_validate(dump_run_state(run)),
created_at=now,
)
record = WorkflowRunRecord(
id=key,
status=status,
resume_readiness=readiness,
environment=environment,
latest_checkpoint_id=checkpoint_id,
created_at=created_at,
updated_at=now,
)
store.save_checkpoint(checkpoint)
store.save_run(record)
return record
def restore_interrupted_run(
store: RunStore, run_id: str
) -> tuple[WorkflowRunRecord, RunState]:
"""Load a persisted interrupted run, persisting a v1 upgrade first.
A pre-budget (v1) checkpoint receives its one-time defaults and is
rewritten as a new v2 interrupted checkpoint under the same run id and
pinned environment *before* the run is returned, so resume dispatch
never runs on unmigrated state and a failed upgrade fails resume before
any handler runs. Ordinary inspection uses :func:`load_stored_run`,
which decodes v1 prospectively without mutating the store.
"""
record = store.get_run(run_id)
if record.status is not StoredRunStatus.INTERRUPTED:
raise ValueError(f"workflow run {run_id!r} is not interrupted")
checkpoint = store.get_latest_checkpoint(run_id)
run, upgraded = load_run_state_with_upgrade(
checkpoint.state.model_dump(mode="json")
)
if upgraded:
record = persist_stopped_run(
store=store,
environment=record.environment,
run=run,
run_id=run_id,
)
return record, run
def load_stored_run(store: RunStore, run_id: str) -> tuple[WorkflowRunRecord, RunState]:
"""Load any stopped run record together with its latest typed checkpoint."""
record = store.get_run(run_id)
if record.status is StoredRunStatus.ADMITTED:
raise ValueError(
f"workflow run {run_id!r} is admitted but has no stopped checkpoint yet"
)
checkpoint = store.get_latest_checkpoint(run_id)
return record, load_run_state(checkpoint.state.model_dump(mode="json"))
def validate_pinned_resume_environment(
*,
record: WorkflowRunRecord,
sources: list[AvailableSource],
) -> list[DependencyDiagnostic]:
"""Revalidate exact stored graph definitions before a resume mutates state."""
environment = record.environment
diagnostics = validate_deployment_dependencies(
artifact=environment.root_artifact,
deployment=environment.deployment,
sources=sources,
)
for child in environment.child_artifacts:
diagnostics.extend(
validate_deployment_dependencies(
artifact=child,
deployment=environment.deployment,
sources=sources,
)
)
return diagnostics
def has_blocking_diagnostics(diagnostics: list[DependencyDiagnostic]) -> bool:
"""Return whether dependency diagnostics prohibit executing a resume."""
return any(item.severity is DiagnosticSeverity.ERROR for item in diagnostics)
def mark_resume_blocked(
*,
store: RunStore,
record: WorkflowRunRecord,
diagnostics: list[DependencyDiagnostic],
) -> WorkflowRunRecord:
"""Record blocked readiness without writing a new execution checkpoint."""
blocked = record.model_copy(
update={
"resume_readiness": ResumeReadiness.BLOCKED,
"diagnostics": diagnostics,
"updated_at": datetime.now(UTC),
}
)
store.save_run(blocked)
return blocked