Set up a full node on Cronos Mainnet

 

Intermediate

 

Set up a full node on Cronos Mainnet

This tutorial is a detailed documentation that explains how to set up a full node on the Cronos mainnet. It begins with a list of pre-requisites that include minimum machine requirements, the supported OS, and other details to be prepared to run a node. The tutorial then moves on to explain how to get the Cronos mainnet binary and configure the cronosd. It also provides instructions on how to clean up the old blockchain data and start over again. The tutorial ends with a detailed upgrade guide for the Huygen network upgrade and how to patch unlucky transactions.

 

Pre-requisites

 

Supported OS

We officially support macOS, Windows, and Linux only. Other platforms may work but there is no guarantee. We will extend our support to other operating systems after we have stabilised our current architecture.

Prepare your machine

To run Cronos Mainnet nodes, you will need a machine with the following minimum requirements to run different types of nodes:

  • Pruned node (setting pruning=everything)

    • Storage: ~25G*

    • RAM: 4 GB (LevelDB) or 64G RAM (RocksDB)***

    • CPU: 4-core

  • Default full node (setting pruning=default)

    • Storage: ~1.5T**

    • RAM: 4 GB (LevelDB) or 64G RAM (RocksDB)***

    • CPU: 4-core

  • Archive node (setting pruning=nothing)

    • Storage: ~2.8T**

    • RAM: 4 GB (LevelDB) or 64G RAM (RocksDB)***

    • CPU: 4-core

*Only in case state-sync enabled. ** e.g. Note that size of snapshots from Quicksync will keep growing. *** Note that during a state-sync the node might require higher RAM than 3GB but, returns to normal after state-sync has finished.

Note that all depends on the type of node you are running and settings will vary depending on your usage.

Mainnet

 

Step 0: Notes on “Huygen” Network upgrade

Before we start, please note that there was “Huygen” network upgrade at the block height 2,693,800, which requires the node operator to update their Cronos Mainnet binary cronosd from v0.6.* to v0.7.0.

For the host who would like to build a Full Node with complete blockchain data from scratch, one would need to:

 
 

Block Height

 

Binary Version

 

Instruction

 

1 ~ 2,693,800

​cronos_v0.6.*​

Start the node with the older binary version cronos_v0.6.* Sync-up with the blockchain until it reaches the target upgrade block height 2,693,800

2,693,800 ~ 3982500

​cronos_v0.7.0​

After it reaches the block height 2,693,800, update app.toml with new config items Update the binary to cronos_v0.7.0 Restart the node

3982500

​cronos_v0.8.3​

After reaching block height, update iavl-disable-fastnodein app.toml Update the binary to cronos_v0.8.3 Restart the node

6542800

​cronos_v1.0.2​

After reaching block height, update app-db-backendin app.toml. Update the binary to cronos_v1.0.2 Restart the node
 

​

Users can refer to the upgrade guide of “Huygen” for the detailed upgrade steps.

To patch “unlucky” transactions, follow this guide on patching unlucky tx​

 

Step 1. Get the Cronos Mainnet binary

Remarks:

The following is the minimal setup for a validator node / full node.

To simplify the following step, we will be using Linux (Intel x86) for illustration. Binaries for Mac (Intel x86 / M1) and Windows are also available.

  • To install released Cronos Mainnet binaries from github:

     
    $ curl -LOJ https://github.com/crypto-org-chain/cronos/releases/download/v0.6.11/cronos_0.6.11_Linux_x86_64.tar.gz
$ tar -zxvf cronos_0.6.11_Linux_x86_64.tar.gz

     

    Afterward, you can check the version of cronosd by

     
    $ ./cronosd version
0.6.11

     

Step 2. Configure cronosd

 

Step 2-0 (Optional) Clean up the old blockchain data

  • If you have joined cronostestnet_338-3 before, you would have to clean up the old blockchain data and start over again, it can be done by running:

    $ ./cronosd unsafe-reset-all

     

    and remove the old genesis file by

    $ rm ~/.cronos/config/genesis.json

     

Before kick-starting your node, we will have to configure your node so that it connects to the Cronos mainnet:

Step 2-1 Initialize cronosd

  • First of all, you can initialize cronosd by:

    $ ./cronosd init [moniker] --chain-id cronosmainnet_25-1

     

    This moniker will be the displayed id of your node when connected to Cronos Chain network.

    When providing the moniker value, make sure you drop the square brackets since they are not needed. The example below shows how to initialize a node named pegasus-node :

    $ ./cronosd init pegasus-node --chain-id cronosmainnet_25-1

     

Note:

  • Depending on your cronosd home setting, the cronosd configuration will be initialized to that home directory. To simply the following steps, we will use the default cronosd home directory ~/.cronos/ for illustration.

  • You can also put the cronosd to your binary path and run it by cronosd

Step 2-2 Configure cronosd

  • Download and replace the Cronos Mainnet genesis.json by:

    $ curl https://raw.githubusercontent.com/crypto-org-chain/cronos-mainnet/master/cronosmainnet_25-1/genesis.json > ~/.cronos/config/genesis.json

     

  • Verify sha256sum checksum of the downloaded genesis.json. You should see OK! if the sha256sum checksum matches.

    $ if [[ $(sha256sum ~/.cronos/config/genesis.json | awk '{print $1}') = "58f17545056267f57a2d95f4c9c00ac1d689a580e220c5d4de96570fbbc832e1" ]]; then echo "OK"; else echo "MISMATCHED"; fi;
OK!

     

NOTE: For Mac environment, sha256sum was not installed by default. In this case, you may setup sha256sum with this command:

function sha256sum() { shasum -a 256 "$@" ; } && export -f sha256sum

 

  • For network configuration, in ~/.cronos/config/config.toml, validator nodes need to modify the configurations of seed, create_empty_blocks_interval and timeout_commit

     
    $ sed -i.bak -E 's#^(seeds[[:space:]]+=[[:space:]]+).*$#\1"0d5cf1394a1cfde28dc8f023567222abc0f47534@cronos-seed-0.crypto.org:26656,3032073adc06d710dd512240281637c1bd0c8a7b@cronos-seed-1.crypto.org:26656,04f43116b4c6c70054d9c2b7485383df5b1ed1da@cronos-seed-2.crypto.org:26656,337377dcda43d79c537d2c4d93ad3b698ce9452e@bd-cronos-mainnet-seed-node-01.bdnodes.net:26656"#' ~/.cronos/config/config.toml
$ sed -i.bak -E 's#^(create_empty_blocks_interval[[:space:]]+=[[:space:]]+).*$#\1"5s"#' ~/.cronos/config/config.toml
$ sed -i.bak -E 's#^(timeout_commit[[:space:]]+=[[:space:]]+).*$#\1"5s"#' ~/.cronos/config/config.toml

     

  • If you would like to build an archive node that allows you to query all the historical block data – kindly update the pruning setting to "nothing" by

 
$ sed -i.bak -E 's#^(pruning[[:space:]]+=[[:space:]]+).*$#\1"nothing"#' ~/.cronos/config/app.toml

 

NOTE: For Mac environment, if jq is missing, you may install it by: brew install jq

 

Step 3. Run everything

CAUTION: This page only shows the minimal setup for validator / full node.

Furthermore, you may want to run full nodes as sentries (see Tendermint), restrict your validator connections to only connect to your full nodes, test secure storage of validator keys etc.

​

Once cronosd has been configured, we are ready to start the ode and sync the blockchain data:

  • Start cronosd, e.g.:

$ ./cronosd start

 

Remarks:

If you see errors saying too many files opened..., then you need to set a higher number for maximum open file descriptors in your OS.

If you are on OSX or Linux, then the following could be useful:

# Check current max fd 
$ ulimit -n 

# Set a new max fd 
$ ulimit -Sn [NEW_MAX_FILE_DESCRIPTOR] 

# Example 
$ ulimit -Sn 4096

 

  • (Optional for Linux) Start cronosd with systemd service, e.g.:

$ curl -s https://raw.githubusercontent.com/crypto-org-chain/cronos-docs/master/systemd/create-service.sh -o create-service.sh && curl -s https://raw.githubusercontent.com/crypto-org-chain/cronos-docs/master/systemd/cronosd.service.template -o cronosd.service.template

$ chmod +x ./create-service.sh && ./create-service.sh

$ sudo systemctl start cronosd

# view log

$ journalctl -u cronosd -f

 

Example: /etc/systemd/system/cronosd.service created by script

# /etc/systemd/system/cronosd.service

[Unit]
Description=cronosd
ConditionPathExists=/usr/local/bin/cronosd
After=network.target


[Service]
Type=simple
User=ubuntu
WorkingDirectory=/usr/local/bin
ExecStart=/usr/local/bin/cronosd start --home /home/ubuntu/.cronos
Restart=on-failure
RestartSec=10
LimitNOFILE=50000


[Install]
WantedBy=multi-user.target

 

It should begin fetching blocks from the other peers.

  • You can query the node syncing status by

    $ ./cronosd status 2>&1 | jq '.SyncInfo.catching_up'

     

    If the above command returns false, It means that your node is fully synced; otherwise, it returns true and implies your node is still catching up.

  • One can check the current block height by querying the public full node by:

    curl -s https://rpc.cronos.org/commit | jq "{height: .result.signed_header.header.height}"

     

    and you can check your node’s progress (in terms of block height) by

    $ ./cronosd status 2>&1 | jq '.SyncInfo.latest_block_height'