hard20 pts · first answer countsobservabilityhistogramssli

Bucket edges and the SLO

The latency SLO is "99% of requests under 300 ms", but the histogram has buckets at 0.25 and 0.5 seconds. What is the problem?

A.Histograms cannot be used for latency SLOsB.The buckets make the SLO 50% stricter automaticallyC.Prometheus rounds 300 ms to 250 ms by designD.You cannot count requests under 300 ms exactly; add a 0.3 bucket

Discussion 0

Join free to post in the discussion.

Bucket edges and the SLO

Discussion 0

Related questions