Hive Developer logo

Hive Developer Portal

Exchange Node

Setting up a node for exchanges.

Intro

By defining a range of accounts to track as a JSON pair ["from", "to"], we can instruct hived to only track the exact accounts that involve the exchange.

These instructions are similar to setting up a Get Transaction Node, but requires far less time to sync because it is configured to focus on only the exchange account(s).

If you prefer a docker version of these instructions, please refer to:

exchangequickstart.md

Sections

Minimum Requirements

This tutorial assumes Ubuntu Server 18.04 LTS 16GB RAM and 320GB SSD/HDD.

Building hived

sudo apt-get update
sudo apt-get dist-upgrade
sudo apt-get install autoconf automake autotools-dev bsdmainutils \
  build-essential cmake doxygen gdb libboost-all-dev libreadline-dev \
  libssl-dev libtool liblz4-tool ncurses-dev pkg-config python3-dev \
  python3-pip nginx fcgiwrap awscli gdb libgflags-dev libsnappy-dev zlib1g-dev \
  libbz2-dev liblz4-dev libzstd-dev
mkdir -p ~/src
cd ~/src
git clone --branch stable https://github.com/steemit/steem.git
cd steem
git submodule update --init --recursive
mkdir -p build
cd build
cmake \
  -DCMAKE_BUILD_TYPE=Release \
  -DLOW_MEMORY_NODE=ON \
  -DCLEAR_VOTES=ON \
  -DSKIP_BY_TX_ID=OFF \
  -DHIVE_LINT_LEVEL=OFF \
  -DENABLE_MIRA=OFF \
  ..
make -j$(nproc)
sudo make install

Configure Node

mkdir -p ~/hive_data
cd ~/hive_data
hived --data-dir=.

At the startup banner, press ^C (Ctrl+C) to exit hived. As a side effect, a default data-dir is created. Now we can purge the empty blockchain and create config.ini as follows:

rm -Rf blockchain
nano config.ini

Then make the following changes to the generated config.ini:

To summarize, the changed values are:

plugin = p2p webserver account_history block_api condenser_api database_api account_history_api
account-history-track-account-range = ["nameofaccount", "nameofaccount"]
shared-file-size = 54G
p2p-endpoint = 0.0.0.0:2000
webserver-http-endpoint = 0.0.0.0:8751
webserver-ws-endpoint = 0.0.0.0:8752

Note that account-history-track-account-range can be a range of accounts, or multiple ranges, if specified more than once (see: account-history-track-account-range).

Save config.ini.

Latest Block Log

Download the block log (optional but recommended).

cd ~/hive_data
mkdir -p blockchain
wget -O blockchain/block_log https://s3.amazonaws.com/steemit-dev-blockchainstate/block_log-latest
hived --data-dir=. --replay-blockchain

Sync Node

If you did not download the latest block log:

cd ~/hive_data
hived --data-dir=. --resync-blockchain

After replay or resync is complete, the console will display Got ## transactions from .... It’s possible to close hived with ^C (Ctrl+C). Then, to start the node again:

cd ~/hive_data
hived --data-dir=.

Troubleshooting

Problem: Got an error while trying to compile hived:

c++: internal compiler error: Killed (program cc1plus)

Solution: Add more memory or enable swap.

To enable swap (do not enable swap on a VPS like Digital Ocean):

sudo dd if=/dev/zero of=/var/swap.img bs=1024k count=4000
sudo chmod 600 /var/swap.img
sudo mkswap /var/swap.img
sudo swapon /var/swap.img

Problem: Got an error while replaying:

IO error: While open a file for appending: /root/hive_data/./blockchain/rocksdb_witness_object/012590.sst: Too many open files

Solution: You’re using MIRA, but this tutorial recommends not to (-DENABLE_MIRA=OFF). If you really intend to try MIRA, you will need to set higher limits. Note, if you are also running hived as root (not recommended), you must explicitly set hard/soft nofile/nproc lines for root instead of * in /etc/security/limits.conf.

To set the open file limit …

sudo nano /etc/security/limits.conf

Add the following lines:

*      hard    nofile     94000
*      soft    nofile     94000
*      hard    nproc      64000
*      soft    nproc      64000

To set the fs.file-max limit …

sudo nano /etc/sysctl.conf

Add the following line:

fs.file-max = 2097152

Load the new settings:

sudo sysctl -p

Once you save these files, you may need to logout and login again.

Back to top