indy-read-replica

Read Replica for Hyperledger Indy Ledgers.

Hyperledger Indy Read Replica is a sort of observer node for Indy Network that holds copies of the ledger transactions and can be used by clients that “trust” the replica.

• The use of read replicas will reduce the read load on the Indy ledgers it replicates.
• Each replica stays up to date with the transactions on the ledgers it connects to.
• Agents using the replica need not connect directly with the ledger.

This project focuses on improving the scalability of read transactions on Hyperledger Indy.

• Allows scaling of read transactions independent of the indy network. I.e support more clients.
• Makes it easy to onboard new developers to Indy. Using a REST interface for indy read allows any client that can make a http request to read from any Hyperledger Indy network.
• Solve network, port and firewall issues caused by reliance on ZMQ by the Indy network.

How it works

Indy read consists of:

• a daemon that reads transactions from all three Indy ledgers (pool, config, domain), and stores them on an elasticsearch index.
• an API server that exposes endpoints analogous to the indy read requests. The server reads from the elasticsearch index.Read more.
• Transactions are stored in Elasticsearch (storage format info)

Structure

- api/        - API for querying db-stored transactions
- daemon/     - process searching for ledger transactions, storing them in a database
- packages/indyscan-storage/    - shared library for app and daemon - how to store and retrieve transactions in db
- packages/indyscan-txtype/     - shared library contaning domain knowledge about indy transactions

Start test instance locally in localhost

The easiest way to get started with indy read would be to use this script. It starts an indy read instance and connects to the candy production netowrk. in order to reduce start up time, the first 1523 transactions are read from here which at the time of writing is all the transactions in the candy production domain ledger.

for Custom deployments or running on host

If:

• you want to deploy indy read replicas on your host machine
• you want to deploy indy read replicas against some particular Indy network
• you want to deploy indy read replicas on intranet

First take a look at the individual services here and here, you will find useful more info about the individual services you need and how to configure them.

This docker-compose will startup an elasticsearch and kibana instance and this other docker-compose will startup both the daemon and api services. You might just need small modifications for your use case.

docker compose -f docker-compose.es.yml -f start_test/docker-compose.dev.yml up -d

Configuration

See Daemon and API server for configuration information.

Usage

You can find out how to setup Indy Read Replica for local tests here.