Skip to content

Latest commit

 

History

History
114 lines (85 loc) · 6.62 KB

How-to-setup-a-Steem-Smart-Contracts-node.md

File metadata and controls

114 lines (85 loc) · 6.62 KB
Raw

1. Environment setup

2. Install the Smart Contracts node

To install the app, simply follow these steps:

  • get the files from the repository:

    • via the git cli: git clone https://github.com/hive-engine/steemsmartcontracts.git

  • cd into the newly created project folder; for Hive Engine, make sure you are on the hive-engine branch (for Steem Engine use the witnesses branch). You may consider using a tagged release as well instead to match the primary nodes, in which case you should checkout a release tag e.g. he_v1.2.0

    • git checkout hive-engine

  • in your console type the following command in the folder that contains the files downloaded from the previous step:

    • npm install

3. Configure the node

The config.json file has all the settings to make sure your node listens to the Hive blockchain and generates the proper genesis block. You shouldn't need to change any of the default config, it can just be used as is. If you are configuring a witness, then additionally set witnessEnabled to true, and copy .env.example to .env to set up your witness settings. Be careful as configuring asks for a witness private signing key in plain text.

(the following file is a pre-configuration that will listen to the sidechain id "mainnet-hive" and will start up a HTTP JSON RPC server listening on port 5000)

{
    "chainId": "mainnet-hive",   // the id of the sidechain that the node will listen to
    "rpcNodePort": 5000,   // port of the JSON RPC server that people will use to retrieve data from your node
    "p2pPort": 5001,
    "databaseURL": "mongodb://localhost:27017",   // url to your MongoDB server
    "databaseName": "hsc",   // name of the MongoDB database
    "blocksLogFilePath": "./blocks.log",   // path to a blocks log file (used with the replay function)
    "javascriptVMTimeout": 10000,   // the timeout that will be applied to the JavaScript virtual machine, needs to be the same on all the nodes of the sidechain
    // array of Hive full nodes used for failover
    "streamNodes": [
        "https://api.openhive.network",
        "https://api.hive.blog",
        "https://anyx.io",
        "https://api.hivekings.com"
    ],
    "startHiveBlock": 41967000,   // last Hive block parsed by the node
    "genesisHiveBlock": 41967000,   // first block that was parsed by the sidechain, needs to be the same on all nodes listening to the sidechain id previously defined
    "witnessEnabled": false
}

4. Start the node

You can easily start the node by typing the following command in the folder where the node was installed:

npm run start

To have it run as a daemon background process, allowing you to close your terminal window without stopping the node, you can use this command instead:

nohup npm start &

You may also use pm2 as well, in which case you should run with the command

pm2 start app.js --no-treekill --kill-timeout 10000 --no-autorestart

Or else pm2 stop will kill all sub processes prematurely and not allow for a clean exit.

5a. Replay from a blocks.log file (DO NOT USE, WILL NOT WORK)

When starting a node for the first time you can either replay the whole sidechain from the Hive blockchain (which can last very long) or replay from a blocks.log file. The blocks.log file is actually the table called "chain" that you can find in your MongoDB database.

  • Find a blocks.log file (ask someone to provide you a JSON version of their "chain" table) or download our official one
  • Start the tool via node app.js --replay file

This command will basically read the file located under "blocksLogFilePath" from the "config.json" file and rebuild the sidechain from the blocks stored in this file.

The latest public blocks.log file is available here in zipped format, file size is about 2.8 GB: https://api.hive-engine.com/blocks_01-19-2021.log.tar.gz

5b. Restore a MongoDB dump (recommended approach)

The fastest way to fire up a node is by restoring a MongoDB dump.

The latest public DB snapshot is available here, file size is about 10 GB: http://api2.hive-engine.com/hsc_20210203_b50993579.archive

When using this snapshot, set startHiveBlock to 50993579 in your config file.

  • Make sure node is stopped
  • Download a dump of the MongoDB database
  • Restore it
    • open mongodb shell by running mongo command
    • show dbs
    • use hsc
    • db.dropDatabase()
    • show dbs // to confirm db has been dropped
    • quit()
    • mongorestore --gzip --archive=hsc_20210203_b50993579.archive
  • Update the "config.json" file with the "startHiveBlock" that matches the dump you just restored
  • Start the node

5c. Restore from hive genesis block.

No extra work needed, the default setting will start syncing from the hive genesis block.

6. Checking that your node works

To verify your node is running properly, you can query data from its API. A getStatus query will show you info on the running software version and latest block processed. Refer to Querying the Engine API for details. In the API URL you should replace https://api.hive-engine.com/rpc with http://<YOUR SERVER IP>:5000

So for example, instead of https://api.hive-engine.com/rpc/blockchain you would use http://<YOUR SERVER IP>:5000/blockchain

7. Enabling and Approving Witness

There is a script called witness_action.js that can help with a few quick actions. See node witness_action.js --help for a list of commands, noting it uses the settings in your .env file to perform the custom json broadcasts.

Witnesses dashboard where you can also approve witnesses is on tribaldex: https://tribaldex.com/witnesses