OpenTelemetry Trace Sampling Language Collocations
Practise the standard verbs for configuring OpenTelemetry trace sampling.
0 / 5 completed
1 / 5
Fill in: 'We ___ a tail-based sampling rule so every trace containing an error is kept, even when the head sampler would otherwise have dropped it.'
We 'configure' a sampling rule — the standard collocation for setting up how traces are selected for storage. The other options aren't idiomatic here.
2 / 5
Fill in: 'Sampling every trace at a flat low rate can ___ the one slow request that actually matters missing from the trace store entirely.'
We say flat sampling will 'leave' a rare important trace missing — the standard, natural collocation for the resulting gap. The other options aren't idiomatic here.
3 / 5
Fill in: 'We ___ the sampling rate down on a high-volume healthy endpoint so storage costs don't scale linearly with traffic that's rarely worth inspecting.'
We 'tune' a rate — the standard, simple collocation for adjusting a configuration parameter. The other options are less idiomatic here.
4 / 5
Fill in: 'We ___ the trace context across every service boundary so a single request's full path stays joined into one trace, not several fragments.'
We 'propagate context' — the standard, established OpenTelemetry collocation for carrying trace identifiers between services. The other options aren't the recognised term here.
5 / 5
Fill in: 'We ___ sampled trace spans in the backend UI to confirm a reported latency spike is really coming from the downstream call we suspect.'
We 'inspect' spans — the standard, simple collocation for examining recorded trace data. The other options are less idiomatic here.