Let a sweep run more than four at a time, and name the run a failure was in
Concurrent runs sat at 4 whatever FLOW_MAX_CASCADES said: that setting bounds cascades, and the run drivers read a hardcoded MAX_PARALLEL nobody could reach. FLOW_MAX_RUNS is the knob they read now, --max-runs/--max-cascades/--max-workers are the same three as flags on serve, and the engine says which numbers it started with — which is the only way to tell that a settings file was read. Events keep the run they happened in. The payload always carried it and the persist path dropped it, so reading one run's failures meant filtering the engine-wide list; a batch run's id reaches those events now too, since a run has no journaled item to name itself by. Also: a provisioner's 0 means "no deadline" rather than "cancel on the next reconcile", and a command that reaches no engine says how to start one.
This commit is contained in:
@@ -125,7 +125,12 @@ class SlurmProvisioner:
|
||||
profiles: list[SlurmProfile],
|
||||
ssh_key: str = "",
|
||||
artifact_url: str = "",
|
||||
#: How long a machine sits idle before it gives itself back; 0 keeps it
|
||||
#: for as long as the job runs, which is what a queue paid for in hours
|
||||
#: of wall time wants.
|
||||
max_idle_s: float = 300.0,
|
||||
#: How long a submitted job may take to attach before it is cancelled;
|
||||
#: 0 waits for as long as the scheduler makes it wait.
|
||||
provision_timeout_s: float = 900.0,
|
||||
events: EventBus | None = None,
|
||||
) -> None:
|
||||
@@ -255,7 +260,10 @@ class SlurmProvisioner:
|
||||
if job.worker and job.worker in attached:
|
||||
# It arrived. Asking again is somebody else's decision.
|
||||
del self._outstanding[name]
|
||||
elif time.monotonic() - job.since > self.provision_timeout_s:
|
||||
elif (
|
||||
self.provision_timeout_s > 0
|
||||
and time.monotonic() - job.since > self.provision_timeout_s
|
||||
):
|
||||
del self._outstanding[name]
|
||||
if job.job_id:
|
||||
expired.append(job.job_id)
|
||||
|
||||
Reference in New Issue
Block a user