These instructions are mostly intended for developers.
If you choose to use these instructions for a production setup, be aware that you will still probably need to do additional configuration for your specific OS, environment, use-case, etc. We do our best here to provide a good starting point, but only proceed if you know what you're doing. Ferritepool does not provide support for custom setups.
See other ways to set up Ferritepool on the main README.
Jump to a section in this doc:
Get the latest fecspace code:
git clone https://github.com/ferritecoin/fecspace
cd fecspace
Check out the latest release:
latestrelease=$(curl -s https://api.github.com/repos/ferritecoin/ferritepool/releases/latest|grep tag_name|head -1|cut -d '"' -f4)
git checkout $latestrelease
Turn on txindex
, enable RPC, and set RPC credentials in ferrite.conf
:
txindex=1
server=1
rpcuser=user
rpcpassword=password
Pick an Electrum Server implementation, configure it, and make sure it's synced.
This step is optional. You can run Ferritepool without configuring an Electrum Server for it, but address lookups will be disabled.
Ferritepool needs MariaDB v10.5 or later. If you already have MySQL installed, make sure to migrate any existing databases before installing MariaDB.
Get MariaDB from your operating system's package manager:
# Debian, Ubuntu, etc.
sudo apt-get install mariadb-server mariadb-client
# macOS
brew install mariadb
mysql.server start
Create a database and grant privileges:
sudo mariadb
MariaDB [(none)]> drop database mempool;
Query OK, 0 rows affected (0.00 sec)
MariaDB [(none)]> create database mempool;
Query OK, 1 row affected (0.00 sec)
MariaDB [(none)]> grant all privileges on mempool.* to 'mempool'@'%' identified by 'mempool';
Query OK, 0 rows affected (0.00 sec)
exit
Make sure to use Node.js 16.10 and npm 7.
node -v
Install dependencies with npm
and build the backend:
cd backend
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.1/install.sh | bash
export NVM_DIR="$([ -z "${XDG_CONFIG_HOME-}" ] && printf %s "${HOME}/.nvm" || printf %s "${XDG_CONFIG_HOME}/nvm")"
[ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh"
source ~/.bashrc # reload shell
nvm install 16
nvm use 16
npm install
npm run build
In the backend folder, make a copy of the sample config file:
cp mempool-config.sample.json mempool-config.json
Edit mempool-config.json
as needed.n
In particular, make sure:
- the correct Ferrite Core RPC credentials are specified in
CORE_RPC
- the correct
BACKEND
is specified inMEMPOOL
:- "electrum" if you're using electrs-ltc or cculianu/Fulcrum- "esplora" if you're using electrs-fec
- "none" if you're not using any Electrum Server
Run the Ferritepool backend:
npm run start
You can also set env var MEMPOOL_CONFIG_FILE
to specify a custom config file location:
MEMPOOL_CONFIG_FILE=/path/to/mempool-config.json npm run start
When it's running, you should see output like this:
Mempool updated in 0.189 seconds
Updating mempool
Mempool updated in 0.096 seconds
Updating mempool
Mempool updated in 0.099 seconds
Updating mempool
Calculated fee for transaction 1 / 10
Calculated fee for transaction 2 / 10
Calculated fee for transaction 3 / 10
Calculated fee for transaction 4 / 10
Calculated fee for transaction 5 / 10
Calculated fee for transaction 6 / 10
Calculated fee for transaction 7 / 10
Calculated fee for transaction 8 / 10
Calculated fee for transaction 9 / 10
Calculated fee for transaction 10 / 10
Mempool updated in 0.243 seconds
Updating mempool
With the backend configured and running, proceed to set up the Ferritepool frontend.
The ferritepool backend is static. TypeScript scripts are compiled into the dist
folder and served through a Node.js web server.
As a result, for development purposes, you may find it helpful to set up backend watchers to avoid the manual shutdown/recompile/restart command-line cycle.
First, install nodemon
and ts-node
:
npm install -g ts-node nodemon
Then, run the watcher:
nodemon src/index.ts --ignore cache/
nodemon
should be in npm's global binary folder. If needed, you can determine where that is with npm -g bin
.
Helpful link: https://gist.github.com/System-Glitch/cb4e87bf1ae3fec9925725bb3ebe223a
Run ferrited on regtest:
ferrited -regtest
Create a new wallet, if needed:
ferrite-cli -regtest createwallet test
Load wallet (this command may take a while if you have lot of UTXOs):
ferrite-cli -regtest loadwallet test
Get a new address:
address=$(ferrite-cli -regtest getnewaddress)
Mine blocks to the previously generated address. You need at least 101 blocks before you can spend. This will take some time to execute (~1 min):
ferrite-cli -regtest generatetoaddress 101 $address
Send 0.1 BTC at 5 lit/vB to another address:
ferrite-cli -named -regtest sendtoaddress address=$(ferrite-cli -regtest getnewaddress) amount=0.1 fee_rate=5
See more example of sendtoaddress
:
ferrite-cli sendtoaddress # will print the help
Mini script to generate random network activity (random TX count with random tx fee-rate). It's slow so don't expect to use this to test mempool spam, except if you let it run for a long time, or maybe with multiple regtest nodes connected to each other.
#!/bin/bash
address=$(ferrite-cli -regtest getnewaddress)
ferrite-cli -regtest generatetoaddress 101 $address
for i in {1..1000000}
do
for y in $(seq 1 "$(jot -r 1 1 1000)")
do
ferrite-cli -regtest -named sendtoaddress address=$address amount=0.01 fee_rate=$(jot -r 1 1 100)
done
ferrite-cli -regtest generatetoaddress 1 $address
sleep 5
done
Generate block at regular interval (every 10 seconds in this example):
watch -n 10 "ferrite-cli -regtest generatetoaddress 1 $address"
By default, mining pools will be not automatically updated regularly (config.MEMPOOL.AUTOMATIC_BLOCK_REINDEXING
is set to false
).
To manually update your mining pools, you can use the --update-pools
command line flag when you run the nodejs backend. For example npm run start --update-pools
. This will trigger the mining pools update and automatically re-index appropriate blocks.
You can enabled the automatic mining pools update by settings config.MEMPOOL.AUTOMATIC_BLOCK_REINDEXING
to true
in your mempool-config.json
.
When a coinbase tag
or coinbase address
change is detected, all blocks tagged to the unknown
mining pools (starting from height 130635) will be deleted from the blocks
table. Additionaly, all blocks which were tagged to the pool which has been updated will also be deleted from the blocks
table. Of course, those blocks will be automatically reindexed.
You can manually force the nodejs backend to drop all data from a specified set of tables for future re-index. This is mostly useful for the mining dashboard and the lightning explorer.
Use the --reindex
command to specify a list of comma separated table which will be truncated at start. Note that a 5 seconds delay will be observed before truncating tables in order to give you a chance to cancel (CTRL+C) in case of misuse of the command.
Usage:
npm run start --reindex=blocks,hashrates
Example output:
Feb 13 14:55:27 [63246] WARN: <lightning> Indexed data for "hashrates" tables will be erased in 5 seconds (using '--reindex')
Feb 13 14:55:32 [63246] NOTICE: <lightning> Table hashrates has been truncated
Reference: mempool/mempool#1269