william3johnson/eosio-failover

An EOSIO failover solution which rotates through potential signing public keys and unregisters when needed.

Best Practices

Your various block producer nodes...

• SHOULD all have unique keys, with each public key loaded into the failover solution
• SHOULD NOT run on the same server as the failover
• SHOULD NOT all be in the same data center

This failover solution...

• SHOULD specify the most reliable API endpoint possible (load balanced API when available)
• SHOULD be integrated with notification methods your organization uses
• SHOULD NOT run in the same server or data center as the API endpoint being used
• SHOULD NOT run in the same server or data center as your block production node
• SHOULD NOT be run with an regproducer_key which is either a active or owner key

Configuration

Establish your configuration file by copying the default configuration and then modifying it.

cp config/default.toml config/local.toml

Edit config/local.toml with the appropriate values for your producer. This file is ignored by git and will retain your settings through upgrades.

• name: Internal name of the application for bunyan logging
• api: EOSIO API (standard) from which to retrieve information.
• rounds_missed_threshold: The number of rounds the script will tolerate before moving to the next key.
• regproducer_key: The private key which can sign both eosio::regproducer and eosio::unregprod.
• producer_account: The producer account to monitor and issue failover commands for.
• producer_permission: The permission name of the private key associated with regproducer_key.
• producer_website: The website address of the producer you'd like to broadcast with eosio::regproducer.
• producer_location: The numeric country code of the producer you'd like to broadcast with eosio::regproducer.
• producer_signing_pubkeys: An array of valid public keys assigned to different production nodes.
• slack (OPTIONAL): Object containing a url of a Slack Webhook to broadcast messages, a channel to specify the target, and a chain to identify which blockchain the messages are in regards to.
• level (OPTIONAL): The output level of the logger.
• out (OPTIONAL): The output method of the logger.

Running

After the configuration has been set and the example values replaced, there are multiple ways to run this service:

nodejs

In order to run this script, you will have to compile the TypeScript and then run the resulting javascript.

make lib
node lib/app.js

You should be able to run this script with any nodejs based process handler or from within a tmux session.

docker + docker-compose

A docker and docker-compose configuration has been provided to make deployment as simple as possible.

If you need to install either, refer to the following guides:

• How To Install and Use Docker on Ubuntu 18.04 (DigitalOcean)
• How To Install Docker Compose on Ubuntu 18.04 (DigitalOcean)

To run the failover script within docker using docker-compose, navigate into the root folder of this repository and run:

docker-compose build
docker-compose up -d

Verify its running:

docker-compose ps

Tailing the logs:

docker-compose logs -f --tail="200"

To stop the script:

docker-compose down

Upgrading

The upgrade process from the latest version on github.com should be simple as:

docker-compose stop
git pull
docker-compose build

Development

To develop with this repository, setup the configuration as shown above and then run:

make dev