You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
perf(planner): collapse topk(sum by ...) into CMS-with-heap #744
Optimize the already-correct query-time implementation of:
topk(k, sum by (...) (metric))
into one CountMinSketchWithHeap precompute.
Intended plan
Use CountMinSketchWithHeap.
Set its aggregated labels to the inner sum by (...) labels.
Use sum weighting/subtype.
Emit no query-time aggregation pipeline for this optimized shape.
Prerequisite
The generic nested-aggregation path must exist first, with sum by (...) (metric) as the planned anchor and query-time topk. It is the correctness baseline for this optimization.
Validation
Benchmark the CMS-with-heap path against that baseline and retain Prometheus-equivalence coverage.
Activity
eval: drop classic templates 7 and 9 (rates by job, top-k by rate) (#210
eval: §6.3 workloads re-planned at 10 samples/s for the end-to-end run (