DRAFTThis page is not published. Only visible in development mode.
Guide
Windowed Aggregation over Live Data
Maintain per-interval rollups (counts, sums, rates) with window inside a deferred view.
Tumbling windows
window tumbling { ... } with { interval: '5s' } buckets rows by time. This guide will build a live request-rate rollup end to end.
Reading windowed results
Querying the view returns one row per bucket. This section will cover bucket boundaries and late data.