stormlog.infer.arrival_report

Per-case arrival accounting for inference reports.

Functions

arrival_lines(arrivals[, latency])	One text-report line for a case whose arrivals are worth showing.
arrival_summary(requests[, window])	Count what was offered, sent and completed, and how late requests left.
latency_from_intended_ms(requests)	End-to-end latency measured from each request's intended arrival.

stormlog.infer.arrival_report.arrival_summary(requests, window=None)

Count what was offered, sent and completed, and how late requests left.

requests are all the measured requests of one case, whatever their outcome, and window is the measured infer.phase_window record when the artifact has one. A closed loop has no offered rate: the server’s speed sets it.

Parameters:

• requests (list[dict[str, Any]])

• window (dict[str, Any] | None)

Return type:

dict[str, Any]

stormlog.infer.arrival_report.latency_from_intended_ms(requests)

End-to-end latency measured from each request’s intended arrival.

It includes any time a request waited before it was sent, which a latency measured from the send leaves out.

Parameters:

requests (list[dict[str, Any]])

Return type:

list[float]

stormlog.infer.arrival_report.arrival_lines(arrivals, latency=None)

One text-report line for a case whose arrivals are worth showing.

An open-loop line also gives the latency measured from each request’s intended arrival, which shows any time it waited before being sent.

Parameters:

• arrivals (Any)

• latency (Any)

Return type:

list[str]