Skip to content

Latest commit

 

History

History
210 lines (150 loc) · 12.1 KB

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

File metadata and controls

210 lines (150 loc) · 12.1 KB
Raw

1. Environment setup

  • Make sure you have a Linux server to run the node on. Medium specs are fine for now: 4 GB RAM, a dual core CPU, and at least 300 GB free disk space should work great. (For 4G RAM, you will need to add swap space, so recommend 8G RAM) Ubuntu is recommended, though other flavors of Linux will probably also work.
  • Install NodeJS and NPM. Node version 18.17.0+ is the minimum version recommended for use (although older versions will likely also work for the time being): https://www.digitalocean.com/community/tutorials/how-to-install-node-js-on-ubuntu-20-04
    • If using Node.JS 17+, be sure to use openssl-legacy-provider as a node arg
  • Install MongoDB. At least version 4.4.3 is required (that's what production nodes are running): https://docs.mongodb.com/v4.4/administration/install-community/, which needs to have replication enabled: https://docs.mongodb.com/manual/tutorial/convert-standalone-to-replica-set/
    • MongoDB 5.x has been observed to work as well
    • To enable replication, you just need to add to the replication config in the mongo config: 
      replication:
  replSetName: "rs0"
      and then after restarting mongo with this config, initiating replication in the mongo shell: > rs.initiate()

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/hivesmartcontracts.git

  • cd into the newly created project folder; for Hive Engine, make sure you are on the main 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_v2.0.0

    • git checkout main

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

    • npm ci

3. Configure the node

Copy the config.example.json file to config.json.

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://127.0.0.1:27017",   // 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.deathwing.me",
        "https://rpc.ecency.com",
        "https://hived.emre.sh",
        "https://api.openhive.network"
    ],
    "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": true,
    "defaultLogLevel": "warn",
    "domain" : "",
    "lightNode": {
        "enabled": false,
        "blocksToKeep": 864000,
        "cleanupInterval": 600000
    },
    "rpcConfig" : {
        "maxLimit" : 1000,
        "maxOffset" : -1,
        "maxBatchLength" : 1,
        "logRequests" : false,
        "allowArbitraryProject" : false,
        "disabledMethods" : {
            "blockchain" : ["getBlockRangeInfo"],
            "contracts" : []
        }
    },
    "rpcWebsockets" : {
        "enabled" : true,
        "port" : 5002
    },
    "hashVerificationNode": false,
    "streamerConfig": {
        "antiForkBufferMaxSize": 2,
        "maxQps": 1,
        "lookaheadBufferSize": 5,
        "useBlockApi": false
    }
}

Optional config settings which you may wish to edit:

domain - if you have a domain name for your node, put the fully qualified domain name here. This information will be returned when status queries are made to your node's API.

lightNode.enabled - if set to true, your node will run in light configuration. A light node does not keep full transaction & block information, which reduces the disk space requirements for running the software (old blocks & transactions are cleaned up periodically). Note that you can't switch from a light node back to running as a full node later on (you would need to restore from a full node snapshot).

lightNode.blocksToKeep - only used when lightNode.enabled = true. Specifies how many recent blocks worth of block data & transactions should be kept by light nodes. Defaults to 30 days worth, assuming a perfect 3 second block time.

lightNode.cleanupInterval - only used when lightNode.enabled = true. Amount of time, in ms (default is 600 seconds / 10 minutes), between internal cleanup activity that ensures the light node database does not grow too large.

rpcConfig.maxLimit - max results to be returned from a rpc call, default is 1000

rpcConfig.maxOffset - maximum allowable offset on a rpc call, with -1 being unlimited, default is -1

rpcConfig.maxBatchLength - max batch size allowed for rpc requests, default is 1

rpcConfig.logRequests - if set to true, all requests to the json-rpc server will be logged, with IP and the body of the request, default is false

rpcConfig.allowArbitraryProject - if set to true, all projections will be allowed, the default behvaior is to only allow projection values to be 1 or 0, specifying wether to return a field or not

rpcConfig.disabledMethods.blockchain - a list of blockchain RPC API methods that are unsupported on this particular node. By default getBlockRangeInfo is disabled, as this is a heavy, processor intensive call that should only be used on private nodes. Note that the getStatus API cannot be disabled this way.

rpcConfig.disabledMethods.contracts - same as above but for the contracts RPC API calls. By default, all these methods are enabled.

rpcWebsockets - see documentation here: hive-engine/hivesmartcontracts#5

hashVerificationNode - string or false, default value is false. Whether to tail a different hive engine API and check for hash match before committing a block to the DB. Example of a valid value to enable this feature would be "https://api.hive-engine.com/rpc/blockchain".

streamerConfig.antiForkBufferMaxSize - integer, default value is 2. The number of blocks to check ahead to confirm same fork before sending a block for processing.

streamerConfig.maxQps - integer, default value is 1. Max queries to send at a time per node in config.

streamerConfig.lookaheadBufferSize - integer, default value is 5. Number of blocks to fetch at a time using all nodes in config. Note that if any node in the config is unresponsive it can cause the block fetching to stall. If you see noisy errors in the log files about being unable to fetch blocks, reducing this setting may help.

streamerConfig.useBlockApi - boolean, default value is false. Whether to use block_api get_block as opposed to condenser_api.get_block (which is deprecated) when fetching blocks from Hive nodes.

For more details about light nodes, see the documentation on the pull request here: hive-engine/steemsmartcontracts#144

4. Start the node

Additional step needed for now - Set your machine time zone to UTC: export TZ=UTC for consistent contract behavior until the contract level bug is resolved.

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, CURRENTLY NOT WORKING)

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)
  • 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.

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 6 GB: hsc_05-22-2021_b54107973.archive. When using this snapshot, set startHiveBlock to 54107973 in your config file.

June 7, 2021 update: sorry, we no longer provide public snapshots due to server bandwidth considerations. We may make them available again in the future, but for now there are some Hive Engine witnesses that provide snapshots as a public service. The latest witness provided snapshots are available here:

https://snap.primersion.com/

https://snap.rishipanthee.com/snapshots/ (mirror of https://snap.primersion.com/)

snapshots for light nodes:

https://snap.primersion.com/light/

https://snap.rishipanthee.com/snapshots/light/ (mirror of https://snap.primersion.com/light/)

  • 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_05-22-2021_b54107973.archive
  • Update the "config.json" file with the "startHiveBlock" that matches the dump you just restored (number after b e.g. 54107973)
  • 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

Moidfy the .env file to have your witness details.

ACTIVE_SIGNING_KEY= Your hive private active key 
ACCOUNT= Your hive username


# Only use one of the following:
# NODE_IP= Server's IP(v4 or v6) if using IP
NODE_DOMAIN= Node domain (exclude protocol, so no http(s)://, for example: example.com)


# If your public ports are different from your node config.js ports (requires NAT/port forwarding)
#RPCNODEPORT= Forwarded RPC NAT port
#P2PPORT= Forwarded P2P NAT port

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 and config.json files to perform the custom json broadcasts.

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