winning-wallet-finder_github/recorder/README.md

RECORDER — the full RTDS firehose tape

Third silo: records everything Polymarket's real-time socket emits — every trade, every order match (the maker side of each fill), every market comment, every crypto price tick — into a durable, downloadable archive. ~8M events/day, ~$5-6/month all-in. No keys, no repo clone (code baked into the image), no shared anything with the trading bots: a recorder crash can never touch trading, a bot deploy can never gap the tape.

Capture design (why it's ~complete)

• Dual sockets to wss://ws-live-data.polymarket.com, both subscribed to activity/* + comments/* + crypto_prices/* (rfq/prices probed dead 2026-07-19). The stream silences per-CONNECTION every ~10 min; a single socket measured 92.9% minute-coverage — the twin covers each gap. Cross-socket dedupe on (tx, asset, side, size, price) for trades, payload-hash for aux; a 15s per-connection stale guard (at 3k+ msg/min, 5s of silence is pathological; false trips cost nothing with a twin).
• Segments: hour-rotated to the volume — rtds_YYYYMMDD_HH.jsonl.gz (trades, stable schema) + aux_… (everything else, raw payloads).
• Preservation (user directive 2026-07-19, STRENGTHENED by Stage 0 2026-07-21): nothing is deleted without a verified second copy. The fold sidecar deletes a raw gz only after its Parquet re-reads with a matching row count; the disk guard deletes Parquet only oldest-first and only files the Mac mirror has ACKED. The recorder's own last-resort guard (95% -> drop oldest, ⚠⚠ TAPE LOSS) still protects the live tape from a full disk.

Stage-0 warehouse (2026-07-21): fold on the box, mirror on the Mac

• fold.py (sidecar in the same machine, capture is PID 1): every 2 min, each closed gz segment becomes /data/parquet/<family>/date=YYYY-MM-DD/<segment>.parquet (zstd), row-parity verified, appended to /data/parquet/manifest.jsonl, then the gz is deleted. The volume IS the warehouse: immutable Parquet any client mirrors incrementally. duckdb capped at 384MB (1GB VM) so a busy-hour fold can never starve capture. ~250-400MB/day of parquet -> months of headroom, months more once mirrored+acked files get pruned.
• sync_tape.py (Mac, every 15 min via com.jaxperro.tape-sync + from daily.sh): pulls new manifest entries over flyctl sftp, row-verifies each file, appends its rows into live/rtds.duckdb's native tables (research keeps native speed — views-over-parquet was rejected: per-asset point queries would crawl), records it in ingested, ACKs the box (/data/parquet/acks/<f>.ok). Tape freshness went from "nightly, if the Mac was awake" to ~15 min whenever awake, and the box no longer depends on the Mac to stay healthy for weeks.
• bootstrap_parquet.py: one-shot export of the pre-fold history (2026-07-17..21, existed only in rtds.duckdb) into the mirror — ran 2026-07-21, parity OK (13.84M trades + 3.07M aux). live/parquet/ is the complete durable layer Stage 1 (MotherDuck/ClickHouse) would consume.

The legacy bulk ingest (fallback)

recorder/ingest.py (sftp-first, base64 fallback, integrity guards) stays as the fallback path for a fold-less recorder; with fold running it finds no gz segments and no-ops. Same idempotence keys (ingested by segment name) as sync_tape, so the two can never double-insert.

Ops

flyctl logs -a wwf-recorder --no-tail          # "tape: N trades/min · aux …"
python3 recorder/ingest.py                     # manual pull any time
flyctl deploy --remote-only -c fly.recorder.toml -a wwf-recorder --ha=false
                                               # ANY code change = image rebuild
flyctl ssh console -a wwf-recorder -C "df -h /data"   # volume headroom

Storage: rtds.duckdb grows ~0.5-0.8GB/day on the Mac (~20GB/mo; 1.4TB free as of 2026-07-19 — revisit around the ~100GB mark: Parquet-spool old months).