# deployer_funder PARTIAL (ppdata, #2362, 14 Oct)
File: research-box /opt/sw/ppdata/deployer_funder.db (copy /opt/ppcopy/deployer_funder.db). Table deployer_funder(pool,deployer,funder1,funder2_root,fund_eth,fund_t,hops,kind,venue).
Rows: 14,656 pools, 6,834 distinct deployers, all with deployer. Funder filled for ~1,300 deployers (see board note for latest). Free RPC 429/500 limits it; paid research RPC worker (datajob dfundp, hourly cap 15k calls) adds ~150 wallets/h.
kind: funded by exchange 977, exchange 706, contract 299, funded by contract 62, max hops 6. root==deployer when funder1 is an exchange (walk stops).
Speed limit: free RPC 429s, ~18 wallets/min total. ETA for the remaining ~6.5k deployers: 5+ h (datajobs fund0-3 and dfunder0-3 still walking; rerun `python3 -I /opt/sw/ppdata/df_merge.py` to refresh, idempotent).
Example: select pool,funder1,funder2_root from deployer_funder where funder1 is not null limit 5;
Not built: the live (30 s) writer. Fully covered runner deployers are enough for a first walk-forward on runners; non-runners are mostly missing, so test flagged-vs-unflagged only inside the filled set or wait for the rest.