EPSS Ingest Perf Harness

Sprint: SPRINT_3410_0001_0001_epss_ingestion_storage (Task EPSS-3410-013A / EPSS-3410-014)

Local Run

Prereqs:

Docker available to Testcontainers
.NET 10 SDK (preview, per repo global.json)

Run (310k rows, default):

dotnet run --project src/Scanner/__Benchmarks/StellaOps.Scanner.Storage.Epss.Perf/StellaOps.Scanner.Storage.Epss.Perf.csproj -c Release -- --rows 310000 --output bench/results/epss-ingest-perf.json

Options:

--rows <int>: dataset rows (default: 310000)
--seed <hex>: 64-bit seed in hex without 0x (default: 5EED20251219)
--model-date <YYYY-MM-DD>: model date (default: today UTC)
--postgres-image <image>: Postgres image (default: postgres:16-alpine)
--output <path>: output JSON path

Output Format

The harness writes a single JSON file:

tool: { name, schema }
dataset: { modelDate, rows, seed, compressedSha256, decompressedSha256, modelVersionTag, publishedDate, compressedBytes, decompressedBytes }
environment: { os, framework, processArchitecture, postgresImage }
timingsMs: { datasetGenerate, containerStart, migrations, writeSnapshot, total }
result: { importRunId, rowCount, distinctCveCount }