Records timing of entry / exit / block inclusion and value / fee information for transactions as they go through the mempool and potential block inclusion lifecycle. Data is stored in a PostgreSQL database for later querying.
This project is in support of an MIT project analyzing the Bitcoin fee market.
- PostgreSQL
- Node.js
- RPC and ZeroMQ access to a full Bitcoin node
Install dependancies:
npm iSet the environment variables described in config/index.js.
node .
Runs the mempool watcher which connects to the ZeroMQ port (see http://bitcoincoredocs.com/zmq.html) and records mempool entry / exit for transactions as well as recording blocks.
Transactions are recorded in the txs table. Inputs to transactions are recorded in the txis table
while outputs are recorded in the txos table. A transaction output can become a transaction input
which would represent a respend and in this case will be recorded in both tables. The txis table
has a spent_in_txid column which records the transaction id of the spending transaction wheras the
txos table contains an amount column.
The ways a transaction is added to the txs table include:
- A new transaction is pushed to the mempool via publication through ZeroMQ
- A new transaction is found in the mempool via dumping the mempool through the RPC
- An existing transaction is referenced by a mempool transaction input
Pushed transactions populate the mempool_entry and mempool_exit columns whereas pulled transactions
populate the mempool_seen_at and mempool_unseen_at columns. Transactions with no mempool_*
information were not seen in the mempool and were added in the process of filling out inputs. Use
mempool_seen_at and mempool_unseen_at primarily. If mempool_entry or mempool_exit is populated,
use that instead because that will be more accurate.
Dump data to CSV files with:
\copy (select * from txs) to 'txs.csv' with csv delimiter ',' header
\copy (select * from txis) to 'txis.csv' with csv delimiter ',' header
\copy (select * from txos) to 'txos.csv' with csv delimiter ',' header
\copy (select * from blocks) to 'blocks2.csv' with csv delimiter ',' header