Skip to content

Mining

Jump to bottom
chetanphirke edited this page May 27, 2021 · 147 revisions

Mining Guide

Contents

For the Chinese version (中文版) of these instructions, please visit HERE. (Credits to #Hash1024)

General Information

Epoch Architecture

Zilliqa Epoch Architecture

At the start of each DS Epoch, all mining candidates will run the Proof-of-Work (Ethash algorithm) cycle for a 60 seconds window in order to compete to join the Zilliqa network.

  • Nodes that fulfilled the DS_POW_DIFFICULTY parameter will qualify to join as DS nodes.
  • Nodes that fulfilled the POW_DIFFICULTY parameter will qualify to join as shard nodes.

There are a total of 100 TX epochs (each 1-2 min) within each DS Epoch (2-3 hrs). Every 100th TX epoch is known as the Vacuous epoch.

The vacuous epoch is solely for:

  • Distributing the coinbase rewards to all nodes.
  • Processing of the upgrade mechanism (as there are no forks in pBFT).
  • Writing of persistent state storage (updating of the nodes’ levelDB).

During a vacuous epoch, the network does not process any transactions.

Proof-of-Work algorithm

Zilliqa is using Ethash for its PoW algorithm. Hence, Zilliqa uses a DAG in its proof-of-work algorithm, that is generated at an incremental rate for each DS epoch. The bootstrap DAG size will be roughly 1.02GB.

Network Difficulty

The bootstrapped shard difficulty level for the mainnet is set at 10. This difficulty level is dynamic and adjusts by +/- 0.125 for every +/- 99 deviation from the target 1810 PoW submissions per DS epoch. Maximum increase per DS epochs is +/- 1.

NOTE: Difficulty level is the log2(Difficulty).

Say if there are 1810 seats available in the network:

  • But there are 1909 PoW submissions, the shard difficulty level will increase by 0.125 for the next DS epoch.
  • But there are 2602 PoW submissions, the shard difficulty level will increase by 1 for the next DS epoch.
  • But there are 1711 PoW submissions, the shard difficulty level will decrease by 0.125 for the next DS epoch.
  • But there are 1018 PoW submissions, the shard difficulty level will decrease by 1 for the next DS epoch.

Reward Mechanism

In the Zilliqa network, rewards are split into:

  • Base rewards [25% of total] for all validating nodes (DS/shard) in the network.
  • Flexible rewards [70% of total] that are based on the amount of valid and accepted (first 2/3 signers within a shard) signatures submitted by a node during a TX epoch while doing the pBFT consensus.

Both base rewards and flexible rewards has the same weightage for both DS and shard nodes. All rewards are consolidated over an entire DS epoch and only distributed during the vacuous epoch. Do note that the last 5% of the rewards are given to the lookup and seed nodes.

Say for example, if there are a total of 2400 nodes in the Zilliqa network and the COINBASE_REWARD is set at 197,244.577625571 ZILs per DS Epoch, the reward distribution will be:

  • For Base rewards: 
    197,244.577625571 * 0.25 / 2400
= 20.5463101693 ZILs per node per DS Epoch
  • For Flexible rewards: (on a first-come-first-serve basis) 
    197,244.577625571 * 0.70 / (2,400 * 2/3 [Successful signers] * 99 [TX blocks])
= 0.87166164354 ZILs per valid and accepted signature

NOTE: Guard nodes by Zilliqa are not rewarded. However, the division of rewards before distribution does include the guard nodes in the count. Hence, there are no "bonus" rewards for non-guard nodes.

Find our your daily mining profitability by making a copy of the Reward Calculator and editing the yellow-highlighted cells.

Hardware requirement

The Zilliqa client is officially supported only on Ubuntu OS.

Please follow the steps HERE if you wish to dual boot Windows and Ubuntu 18.04.

If you wish to mine using mining rigs that operate on Windows OS, please follow the remote mining with docker image guide HERE.

Both AMD (with OpenCL) and Nvidia (with OpenCL or CUDA) GPUs are supported for PoW.

The minimum requirements for Zilliqa mining nodes are:

  • x64 Linux operating system (e.g Ubuntu 18.04.5)
  • Recent dual core processor @ 2.2 GHZ. Examples:
    • Intel Core i5 or i7 (Skylake)
    • Intel Xeon (Skylake)
    • AMD Ryzen
  • 4GB DRR3 RAM or higher
  • NAT environment OR Public IP address
  • Any GPUs with at least 2 GB vRAM

Initial Setup

Network setup

NOTE: If you are using a home router, you are most probably in a NAT environment.

If you are in NAT environment, you can either:

  • Do single port forwarding using Option 1a. This should be your DEFAULT OPTION.
  • Enable UPnP mode using Option 1b if your router does support UPnP.

If you have a public IP address, you can skip this network setup entirely.

  • (Option 1a) Port forward to port 33133 for both external port (port range) and internal port (local port). You will also have to select the option for BOTH TCP and UDP protocol in your router menu when port forwarding.

    An example of this process can be found HERE. After port forwarding, you may check if you have successfully port forwarded with this Open Port Check Tool.

  • (Option 1b) Enable UPnP mode on your home router. Please Google how to access your home router setting to enable UPnP, an example can be found HERE. You can check if you have successfully enabled UPnP by installing the following tool:

    sudo apt-get install miniupnpc

    Then type the following in the command line:

    upnpc -s

    You should get a message showing either:

    • "List of UPNP devices found on the network : ..."
    • OR "No IGD UPnP Device found on the network !".

    The first message means UPnP mode has been enabled successfully, while the latter means the enabling of UPnP mode has failed. If you receive the latter message, proceed with using Option 1a instead.

OpenCL driver setup (for AMD/Nvidia)

If you wish to use OpenCL supported GPUs for PoW, please run the following to install the OpenCL developer package:

sudo apt install ocl-icd-opencl-dev

You may need to reboot your PC for the installation to take effect. After reboot, check if your drivers are installed properly with the following command:

clinfo

CUDA driver setup (for Nvidia)

If you wish to use CUDA supported GPU for PoW, please download and install CUDA package from the NVIDIA official webpage. You may need to reboot your PC for the installation to take effect.

Multiple GPUs setup

If you wish to run multiple AMD or Nvidia GPUs concurrently, edit the GPU_TO_USE parameter in the constants.xml file located in your join folder.

The index start from 0 and you can input one or more multiple GPUs by separating their indexes with a ,.

For example:

  • 0 for just 1 GPU.
  • 0, 1, 2 or 0, 2, 4 for 3 GPUs.

Do note that the largest index must correspond to the number of GPUs you have physically in your mining rig.

Mining with docker locally (For CPU or Nvidia GPUs only)

  1. Install Ubuntu 18.04.5 OS by following instructions HERE.

  2. Install Docker CE for Ubuntu by following instructions HERE.

  3. Make a new directory in your Desktop and change directory to it:

    cd ~/Desktop && mkdir join && cd join

  4. Get the joining configuration files:

    wget https://mainnet-join.zilliqa.com/configuration.tar.gz
tar zxvf configuration.tar.gz

  5. Find out your current IP address in the command prompt and record it down.

    NOTE: If you are using Option 1b as stated in the Network Setup above, you can skip this step.

    curl https://ipinfo.io/ip

  6. Run the shell script in your command prompt to launch your docker image.

    • (Option 1) For mining with CPU, launch your docker container:

      ./launch_docker.sh

    • (Option 2) For mining with Nvidia GPUs, please first install the nvidia-docker HERE. Then, change the following parameter in the constants.xml file within the join folder to true:

      <CUDA_GPU_MINE>true</CUDA_GPU_MINE>

      After which, launch your docker container:

      ./launch_docker.sh --cuda

      NOTE: If you wish to run multiple Nvidia GPUs concurrently, you will need to modify your constants.xml file following instructions as found above HERE.

  7. You will be prompted to enter some information as shown below:

    • Assign a name to your container (default: zilliqa):
      [Press Enter to skip if using default]

    • Enter your IP address ('NAT' or *.*.*.*):
      [Key in your IP address as found in step 6 OR NAT if you using Option 1b]

    • Enter your listening port (default: 33133):
      [Press Enter to skip if using default]

      NOTE: DO NOT duplicate your IP address and use different ports on different CPU nodes. You will be blacklisted by the network and hence not be able to receive any flexible rewards.

  8. You are now a miner in the Zilliqa mainnet. You can monitor your progress using:

    tail -f zilliqa-00001-log.txt

  9. To check your locally generated public and private key pairs, you can enter the following in your command prompt:

    less mykey.txt

    The first hex string is your public key, and the second hex string is your private key.

    NOTE: This key pair is generated locally on your disk. Do remember to keep your private key somewhere safe!

  10. To check your balance, get your address and input the address in the search bar of https://viewblock.io/zilliqa:

    less myaddr.txt

  11. To stop the mining client, stop your running docker container:

    sudo docker ps (get zilliqa's CONTAINER_ID here)
sudo docker stop zilliqa
sudo docker rm CONTAINER_ID

Mining with docker remotely

The setup architecture is illustrated in the image shown below. All communications between these two parties is via JSON-RPC protocol.

1-to-many

  • The CPU node instance will run the Zilliqa client and carry out the pBFT consensus process to receive rewards.
  • The GPU rigs in the GPU cluster will run Zilminer on a separate GPU cluster to do PoW mining and provide PoW solutions directly to the CPU node.

For hooking up several GPU rigs in the GPU cluster to a single CPU node, you will be required to do the following steps:

  1. Create a single local/remote CPU node instance with Ubuntu 18.04 OS installed following instructions HERE.

  2. Install Docker CE for Ubuntu on your CPU node instance by following instructions HERE.

  3. Make a new directory in your Desktop and change directory to it:

    cd ~/Desktop && mkdir join && cd join

  4. Get the joining configuration files:

    wget https://mainnet-join.zilliqa.com/configuration.tar.gz
tar zxvf configuration.tar.gz

  5. Find out your current IP address in the command prompt and record it down:

    curl https://ipinfo.io/ip

  6. Edit your constant.xml file in your configuration folder:

    • Set GETWORK_SERVER_MINE to true.

    • Set GETWORK_SERVER_PORT to the port you will be using to GetWork. (default is 4202)

    • Set the following mining parameters to false:

      <CUDA_GPU_MINE>false</CUDA_GPU_MINE>
<FULL_DATASET_MINE>false</FULL_DATASET_MINE>
<OPENCL_GPU_MINE>false</OPENCL_GPU_MINE>
<REMOTE_MINE>false</REMOTE_MINE>

  7. Run the shell script in your command prompt to launch your docker image.

    ./launch_docker.sh

  8. You will be prompted to enter some information as shown below:

    • Assign a name to your container (default: zilliqa):
      [Press Enter to skip if using default]

    • Enter your IP address ('NAT' or *.*.*.*):
      [Key in your IP address as found in step 5 OR NAT if you chose Option 1b during Network setup]

    • Enter your listening port (default: 33133):
      [Press Enter to skip if using default]

      NOTE: DO NOT duplicate your IP address and use different ports on different CPU nodes. You will be blacklisted by the network and hence not be able to receive any flexible rewards.

  9. Once the CPU Zilliqa node is running, you can install Zilminer on your separate GPU rigs:

  10. Setup your Zilminer on your separate GPU rigs with the following command:

    zilminer -P zil://wallet_address.worker_name@zil_node_ip:get_work_port

    NOTE: You have to change the wallet_address, worker_name, zil_node_ip, and get_work_port accordingly.

    • For wallet_address : You can input any arbitrary Zilliqa address if you are mining solo.
    • For worker_name : You can input any arbitrary worker name you desire.
    • For zil_node_ip : Please input the IP address of the Zilliqa node in step 5.
    • For get_work_port : Please input the port used in GETWORK_SERVER_PORT. Default is 4202.

  11. You are now a proxy miner in the Zilliqa mainnet. You can monitor your progress on your CPU node using:

    tail -f zilliqa-00001-log.txt

  12. To check your locally generated public and private key pairs, you can enter the following in your command prompt on your CPU node:

    less mykey.txt

    The first hex string is your public key, and the second hex string is your private key.

    NOTE: This key pair is generated locally on your disk. Do remember to keep your private key somewhere safe!

  13. To check your balance, get your address and input the address in the search bar of https://viewblock.io/zilliqa:

    less myaddr.txt

  14. To stop the mining client, stop your running docker container on the CPU node and kill your Zilminer process on your GPU rigs:

    sudo docker stop zilliqa

Mining with native build locally

  1. Make a new directory for Zilliqa client:

    cd ~/Desktop && mkdir Zilliqa

  2. Make a new directory for Scilla binary:

    mkdir Scilla

  3. Make a new directory for the join folder:

    mkdir join

  4. Clone the Scilla repository and change directory to it:

    git clone https://github.com/Zilliqa/Scilla.git Scilla && cd Scilla && git checkout v0.9.1

  5. Find out your Scilla directory path and record it down:

    pwd

  6. First, download the Scilla binary's dependencies for Ubuntu following instructions found HERE.

  7. Then, build the Scilla binary:

    make clean; make

  8. Clone the Zilliqa repository and change directory to it:

    cd ~/Desktop && git clone https://github.com/Zilliqa/Zilliqa.git Zilliqa && cd Zilliqa && git checkout v8.0.4

  9. Find out your Zilliqa directory path again and write it down:

    pwd

  10. First, download the Zilliqa client's dependencies. Then, build Zilliqa with Option 1 for CPU mining, or with Option 2/Option 3 for GPU mining.

    • Download the dependencies:

      sudo apt-get update
sudo apt-get install git libboost-system-dev libboost-filesystem-dev libboost-test-dev \
libssl-dev libleveldb-dev libjsoncpp-dev libsnappy-dev libmicrohttpd-dev \
libjsonrpccpp-dev build-essential pkg-config libevent-dev libminiupnpc-dev \
libprotobuf-dev protobuf-compiler libcurl4-openssl-dev libboost-program-options-dev \
libssl-dev python3-dev libboost-python-dev python3-setuptools python3-pip gawk

    • Run the following to install latest version of cmake.We suggest to install cmake 3.19 or any version >=3.16:

      wget https://github.com/Kitware/CMake/releases/download/v3.19.3/cmake-3.19.3-Linux-x86_64.sh
mkdir -p "${HOME}"/.local
bash ./cmake-3.19.3-Linux-x86_64.sh --skip-license --prefix="${HOME}"/.local/
export PATH=$HOME/.local/bin:$PATH
cmake --version
rm cmake-3.19.3-Linux-x86_64.sh

    • (Option 1) Build Zilliqa for CPU mining:

      ./build.sh

    • (Option 2) Build Zilliqa for Nvidia GPU mining with CUDA support:

      ./build.sh cuda

    • (Option 3) Build Zilliqa for AMD or Nvidia GPU mining with OpenCL support:

      ./build.sh opencl

  11. Download and unpack the compressed joining configuration file:

    cd ../join && wget https://mainnet-join.zilliqa.com/configuration.tar.gz && tar zxvf configuration.tar.gz

  12. Edit the constants.xml in your join folder to key in the Scilla directory for the SCILLA_ROOT parameter. An example is shown below:

    <SCILLA_ROOT>/home/ubuntu/Scilla</SCILLA_ROOT>

  13. (Optional) If you wish to mine with GPUs, please continue to edit the following parameters in the constants.xml file in your join folder:

    NOTE: If you wish to run multiple GPUs concurrently, you will need to modify your constants.xml file following instructions as found above HERE.

    • For AMD GPUs: Change FULL_DATASET_MINE parameter from false to true. Change OPENCL_GPU_MINE parameter from false to true.
    • For Nvidia GPUs: Change FULL_DATASET_MINE parameter from false to true. Change CUDA_GPU_MINE parameter from false to true.

  14. Find out your current IP address in the command prompt and record it down.

    NOTE: If you are using Option 1b as stated in the Network Setup above, you can skip this step.

    curl https://ipinfo.io/ip

  15. Install python dependencies:

    sudo apt install python-pip
export LC_ALL=C
pip install request requests clint futures

  16. Launch the Zilliqa client:

    ./launch.sh

  17. You will be prompted to key in the following details:

    • Enter the full path of your zilliqa source code directory:
      [Key in the path you found it step 8]

    • Enter your IP address (NAT or *.*.*.*):
      [Key in your IP address as found in step 13 OR NAT if you are using Option 1b]

    • Enter your listening port (default: 33133):
      [Press Enter to skip if using default]

      NOTE: DO NOT duplicate your IP address and use different ports on different CPU nodes. You will be blacklisted by the network and hence not be able to receive any flexible rewards.

  18. You are now a miner in Zilliqa mainnet. You can monitor your progress using:

    tail -f zilliqa-00001-log.txt

  19. To check your locally generated public and private key pairs, you can enter this in your command prompt:

    less mykey.txt

    The first hex string is your public key, and the second hex string is your private key.

    NOTE: The key pair is generated locally on your disk. Do remember to keep your private key somewhere safe!

  20. To check your balance, get your address and input the address in the search bar of https://viewblock.io/zilliqa:

    less myaddr.txt

  21. To stop Zilliqa client:

    pkill zilliqa

Discussion channels

Channels

Join our official Discord channel: https://discord.gg/XMRE9tt Join the community-managed Telegram channel: https://t.me/zilliqaminer

Want to contribute to this Wiki?

Fork it and send a pull request.

Clone this wiki locally