Chainalytic Framework

Note	THIS IS DEPRECATED, for a better version, please look at https://github.com/yudus-lab/chainalytic-icon

Modular framework for blockchain time-series data and arbitrary chain-state derivation/aggregation, which help to generate more insights of generic blockchains.

Detailed introduction and the story of Chainalytic

How to use pre-built production Docker image

Jump to Provider service API reference

The Purpose

Note	This is not* the tool that simply extracts block/txs data and dumps to other databases (e.g. SQL, NoSQL, BigQuery … )*

There is a huge amount of data deep inside blockchains which could bring potential valuable insights.

This framework focuses on

Derive implicit blockchain data using transactions as input ( transition functions )
Realtime/offline time-series data derivation/aggregation
Maintain independent and diverse chain-states

It helps to easily extract, transform, aggregate and generate new data using your own rule in a modular style.

It targets developers/analysts who are interested in building application or research on

Blockchain analytic and insights
Monitoring tools
Data science, machine learning and AI
Algo trading

I am also a big fan of ICON, thus it comes with one built-in data aggregator for ICON public mainnet.

Architecture

Chainalytic services

Upstream, service ID = 0

Load and pre-processing original block data, then pass it to Aggregator

Aggregator, service ID = 1

Transform and aggregate data

Warehouse, service ID = 2

Take care of all storage tasks

Provider, service ID = 3

Communicate with Warehouse and ask for aggregated data, then provide it to external applications via aiohttp JSON-RPC APIs

Chainalytic Console

Handle data aggregation monitoring.

Quick start

Go to development guide here

Prerequisite

Recommend Ubuntu 18.04 and Python 3.7

Install dependencies

sudo apt update
sudo apt install libsecp256k1-dev libleveldb-dev pkg-config build-essential

python37 -m venv venv
venv/bin/pip install -e .

1. Chainalytic Console

1A. You must setup user config first, run below command to generate it

venv/bin/python launch.py --init-config

Then, in current working dir, there should be 2 config files

.chainalytic/cfg/chain_registry.yml
.chainalytic/cfg/setting.yml

In chain_registry.yml, for ICON public chain, modify it like below

zones:
  - zone_id: 'public-icon'
    zone_name: 'Public ICON mainnet'
    client_endpoint: 'localhost:9000' # Your citizen node endpoint
    chain_db_dir: 'path_to/data/mainnet/.storage/db_<YOUR_IP>:7100_icon_dex'
    score_db_icondex_dir: 'path_to/data/mainnet/.score_data/db/icon_dex'
    direct_db_access: 0 # Set to 1 for fast aggregation
    transforms:
      - stake_history
      - stake_top100
      - recent_stake_wallets
      - abstention_stake
      - funded_wallets
      - passive_stake_wallets

1B. Basic usages

Run all services

venv/bin/python launch.py

Monitor one specific data transform

venv/bin/python launch.py m <TRANSFORM_ID>

Default TRANSFORM_ID is stake_history

List of TRANSFORM_ID:

stake_history
stake_top100
recent_stake_wallets
abstention_stake

Terminate all services

venv/bin/python launch.py stop

Show help

venv/bin/python launch.py -h

2. Chainalytic services

Run individual services ( for debugging )

venv/bin/python -m chainalytic.upstream --endpoint localhost:5500 --zone_id public-icon --working_dir .
venv/bin/python -m chainalytic.aggregator --endpoint localhost:5510 --zone_id public-icon --working_dir .
venv/bin/python -m chainalytic.warehouse --endpoint localhost:5520 --zone_id public-icon --working_dir .
venv/bin/python -m chainalytic.provider --endpoint localhost:5530 --zone_id public-icon --working_dir .

yudus-labs/chainalytic-framework