tt_dispatch_dispatch_waiting_on_upstream

<< Home | << Metrics

Name

Prometheus Metric Name

tt_dispatch_dispatch_waiting_on_upstream

Metric Path (tt-telemetry)

Schema:

{hostname}/tray{tray}/chip{chip}/cq{cq}/tt_dispatch_dispatch_waiting_on_upstream

Example path:

bh-glx-c09u02/tray1/chip2/cq0/tt_dispatch_dispatch_waiting_on_upstream

Description

A point-in-time flag indicating whether, at the moment of the telemetry read, the dispatch core for this command queue (CQ) was stalled waiting on its upstream. In the fast-dispatch pipeline (host → prefetch core → dispatch core → worker cores) the dispatch core’s upstream is the prefetch core, so a true value means the dispatcher has consumed everything the prefetcher handed it and is waiting for more commands.

This points one stage further down the pipeline than tt_dispatch_prefetch_waiting_on_upstream: the prefetcher is not keeping the dispatcher fed. That can be because the prefetcher is itself starved by the host, or because prefetching is the bottleneck. Because it is sampled instantaneously once per interval, pair it with tt_dispatch_dispatch_blocked_count to gauge how often the stall occurs.

A separate metric is emitted for every command queue the chip can support (cq0, cq1, …). CQs absent from a given read, and chips whose firmware bundle is below version 19.12.0, report the default value of 0.

This is an integer flag (only ever 0 or 1) rather than a boolean so that it does not contribute to the overall health status.

Values

Type: Integer

Units: None

Allowable values:

  • 1: The dispatch core was waiting on the prefetch core for more work at read time.

  • 0: The dispatch core was not waiting (also reported when the CQ was absent from the read or dispatch telemetry was unavailable).

Prometheus Labels

Label Name

Value

hostname

The host from which the metric was collected.

hall

The datacenter hall where the host is located. Sourced from the Factory System Descriptor (FSD).

aisle

The datacenter aisle where the host is located. Sourced from the Factory System Descriptor (FSD).

rack

The rack number where the host is located. Sourced from the Factory System Descriptor (FSD).

shelf_u

The shelf U position in the rack where the host is located. Sourced from the Factory System Descriptor (FSD).

tray

The tray (UBB) that the device is located on.

chip

The ASIC location within the tray.

cq

The hardware command queue index (0, 1, …).