pub struct JobRun {Show 14 fields
pub lease_id: Option<LeaseId>,
pub session: Option<SessionKey>,
pub end_cause: Option<EndCause>,
pub fingerprint: String,
pub command: String,
pub cwd: String,
pub declared: JobProfile,
pub peak: ResourceUsage,
pub avg: ResourceUsage,
pub busy: Option<BusyUsage>,
pub started_at_ms: i64,
pub ended_at_ms: i64,
pub noise: Option<NoiseReport>,
pub overlapped: bool,
}Expand description
One finished lease with what its holder’s session used, kept for history-based estimates.
Fields§
§lease_id: Option<LeaseId>Lease the job ran under, if the lease is still stored.
session: Option<SessionKey>Session that ran the job; None for runs recorded before sessions were.
end_cause: Option<EndCause>Why the run ended; None for runs recorded before causes were.
fingerprint: StringNormalized command plus working directory, used to find similar runs.
command: StringThe lease’s label, usually the command as it was run.
cwd: StringWorking directory of the command.
declared: JobProfileProfile the lease reserved, which may already include an earlier estimate.
peak: ResourceUsagePeak usage of the holder’s session during the run.
avg: ResourceUsageAverage usage of the holder’s session during the run.
busy: Option<BusyUsage>Average load while the session was busy; None when it never was, and for runs
recorded before this was measured.
started_at_ms: i64Start of the run, in milliseconds since the Unix epoch.
ended_at_ms: i64End of the run, in milliseconds since the Unix epoch.
noise: Option<NoiseReport>Background load during the run; recorded only for benchmark leases.
overlapped: boolWhether the session held another active lease during the run. Usage is measured per session, so such a run mixes several jobs and estimates leave it out.