Skip to content

Latest commit

 

History

History

galileo

README.md

IoT Service Blockchain Benchmark Suite

Benchmarking IoT Service Blockchain smart contracts using Hyperledger Caliper.

Testbed Setup

Name CPU (Cores) Memory (MB) Disk (GB) IP OS
manager.caliper.example.com 2 2048 40 172.16.0.101 Ubuntu 20.04.4
worker1.caliper.example.com 2 2048 40 172.16.0.102 Ubuntu 20.04.4
worker2.caliper.example.com 2 2048 40 172.16.0.103 Ubuntu 20.04.4
worker3.caliper.example.com 2 2048 40 172.16.0.104 Ubuntu 20.04.4
worker4.caliper.example.com 2 2048 40 172.16.0.105 Ubuntu 20.04.4
worker5.caliper.example.com 2 2048 40 172.16.0.106 Ubuntu 20.04.4

Installation & Running

Prerequisites

  • Install Node.js of version >= 14.13.1 and < 15 on all Caliper nodes.

  • Install make and other build tools (build-essential package on Ubuntu) on all Caliper nodes.

  • Install Eclipse Mosquitto 2.0 on the Caliper manager node (manager.caliper.example.com).

  • Ensure every Caliper node (*.caliper.example.com) can connect to each other and all Hyperledger Fabric nodes (orderer*.example.com, peer*.org*.example.com). A simple way is to achieve this is editting /etc/hosts:

    $ echo "172.16.0.2 orderer1.example.com" >> /etc/hosts
$ echo "172.16.0.3 orderer2.example.com" >> /etc/hosts
$ echo "172.16.0.4 orderer3.example.com" >> /etc/hosts
$ echo "172.16.0.5 peer1.org1.example.com" >> /etc/hosts
$ echo "172.16.0.6 peer2.org1.example.com" >> /etc/hosts
$ echo "172.16.0.7 peer1.org2.example.com" >> /etc/hosts
$ echo "172.16.0.8 peer2.org2.example.com" >> /etc/hosts

$ echo "172.16.0.101 manager.caliper.example.com" >> /etc/hosts
$ echo "172.16.0.102 worker1.caliper.example.com" >> /etc/hosts
$ echo "172.16.0.103 worker2.caliper.example.com" >> /etc/hosts
$ echo "172.16.0.104 worker3.caliper.example.com" >> /etc/hosts
$ echo "172.16.0.105 worker4.caliper.example.com" >> /etc/hosts
$ echo "172.16.0.106 worker5.caliper.example.com" >> /etc/hosts

  • Make sure incoming traffic to port 1833 of the Caliper manager node (manager.caliper.example.com) and port 9100 of all Hyperledger Fabric nodes (orderer*.example.com and peer*.org*.example.com) is allowed by the network firewall. You may optionally allow port 9090 of the Caliper manager node (manager.caliper.example.com) as well to access the Prometheus web portal.

Set up benchmarks

  1. Set up a multi-node Hyperledger Fabric testbed as instructed by Gaia.

  2. Configure Mosquitto to listen on port 1883 and allow anonymous users on the Caliper manager node (manager.caliper.example.com).

    # /etc/mosquitto/conf.d/default.conf
allow_anonymous true
listener 1883

    Then, start the mosquitto systemd service:

    (manager.caliper.example.com) $ sudo systemctl restart mosquitto

  3. Clone this repository to all Caliper nodes (*.caliper.example.com):

    $ git clone https://github.com/nexus-lab/iot-service-blockchain-samples.git

  4. Copy iot-service-blockchain-samples/gaia/crypto-config/ generated when setting up the testbed to all Caliper nodes (*.caliper.example.com) to the same location (iot-service-blockchain-samples/gaia/crypto-config/).

  5. Change working directory to iot-service-blockchain-samples/galileo on all Caliper nodes (*.caliper.example.com).

    $ cd iot-service-blockchain-samples/galileo

  6. Install Node.js dependencies on all Caliper nodes (*.caliper.example.com). You need to authenticating to GitHub Packages to install @nexus-lab/iot-service-blockchain. Also, set up the .npmrc file following Installing a package to let your package manager search for @nexus-lab/iot-service-blockchain on the GitHub Package repository.

    $ npm install

  7. Edit connectors/connection-org1.yaml and connectors/connection-org1.yaml, replace the pem fields with corresponding file content of gaia/crypto-config/peerOrganizations/org*.example.com/peers/peer*.org*.example.com/tls/ca.crt:

    # connectors/connection-org1.yaml
---
...
peers:
  peer1.org1.example.com:
    ...
    tlsCACerts:
      pem: |
        # replace with file content of gaia/crypto-config/peerOrganizations/org1.example.com/peers/peer1.org1.example.com/tls/ca.crt
    ...
  peer2.org1.example.com:
    ...
    tlsCACerts:
      pem: |
        # replace with file content of gaia/crypto-config/peerOrganizations/org1.example.com/peers/peer2.org1.example.com/tls/ca.crt
    ...
...
    # connectors/connection-org2.yaml
---
...
peers:
  peer1.org2.example.com:
    ...
    tlsCACerts:
      pem: |
        # replace with file content of gaia/crypto-config/peerOrganizations/org2.example.com/peers/peer1.org2.example.com/tls/ca.crt
    ...
  peer2.org2.example.com:
    ...
    tlsCACerts:
      pem: |
        # replace with file content of gaia/crypto-config/peerOrganizations/org2.example.com/peers/peer2.org2.example.com/tls/ca.crt
    ...
...

Set up system resource monitors

  1. Download, extract, and install Prometheus node exporter to all Hyperledger Fabric nodes (orderer*.example.com and peer*.org*.example.com):

    ([orderer*|peer*.org*].example.com) $ curl -L https://github.com/prometheus/node_exporter/releases/download/v1.3.1/node_exporter-1.3.1.linux-amd64.tar.gz \
                                           -o /tmp/node_exporter-1.3.1.linux-amd64.tar.gz
([orderer*|peer*.org*].example.com) $ tar -xvf /tmp/node_exporter-1.3.1.linux-amd64.tar.gz -C /tmp
([orderer*|peer*.org*].example.com) $ sudo install \
                                           -o root \
                                           -g root /tmp/node_exporter-1.3.1.linux-amd64/node_exporter /usr/local/bin

  2. Start the node exporter on all Hyperledger Fabric nodes (orderer*.example.com and peer*.org*.example.com) in separate terminals or screens:

    ([orderer*|peer*.org*].example.com) $ /usr/local/bin/node_exporter

  3. Download, extract, and install Prometheus to the Caliper manager node (manager.caliper.example.com).

    (manager.caliper.example.com) $ curl -L https://github.com/prometheus/prometheus/releases/download/v2.34.0/prometheus-2.34.0.linux-amd64.tar.gz \
                                     -o /tmp/prometheus-2.34.0.linux-amd64.tar.gz
([orderer*|peer*.org*].example.com) $ tar -xvf /tmp/prometheus-2.34.0.linux-amd64.tar.gz -C /tmp
([orderer*|peer*.org*].example.com) $ sudo install \
                                           -o root \
                                           -g root /tmp/prometheus-2.34.0.linux-amd64/prometheus /usr/local/bin

  4. On the Caliper manager node (manager.caliper.example.com), change working directory to galileo and start Prometheus in a separate terminal or screen:

    (manager.caliper.example.com) $ /usr/local/bin/prometheus --config.file=./prometheus.yaml

Run Benchmarks

  1. First, start Caliper manager on the Caliper manager node (manager.caliper.example.com):

    (manager.caliper.example.com) $ npm run manager -- \
                                    --caliper-networkconfig connectors/manager.yaml \
                                    --caliper-benchconfig <benchmark config file>

    Replace the <benchmark config file> with the relative path to the *.yaml file under benchmarks/ directory. You may choose benchmark configuration file from either the full or lite subdirectory. The full benchmarks will spawn 2000 connections in total from 5 Caliper workers to the Hyperledger Fabric peers, while the lite benchmarks will only create 10 connections.

    Because some benchmarks depend on the artifacts generated by other benchmarks, it is recommended to run the benchmarks in the following order:

    • (full|lite)/device-registry/register.yaml
    • (full|lite)/device-registry/get.yaml
    • (full|lite)/device-registry/get-all.yaml
    • (full|lite)/service-registry/register.yaml
    • (full|lite)/service-registry/get.yaml
    • (full|lite)/service-registry/get-all.yaml
    • (full|lite)/service-broker/request.yaml
    • (full|lite)/service-broker/respond.yaml
    • (full|lite)/service-broker/get.yaml
    • (full|lite)/service-broker/get-all.yaml
    • (full|lite)/service-broker/remove.yaml
    • (full|lite)/service-registry/deregister.yaml
    • (full|lite)/device-registry/deregister.yaml

  2. Then, start one Caliper worker on each Caliper worker node (worker*.caliper.example.com):

    (worker1.caliper.example.com) $ npm run worker -- \
                                    --caliper-networkconfig connectors/(full|lite)/worker1.yaml \
                                    --caliper-benchconfig <benchmark config file>
(worker2.caliper.example.com) $ npm run worker -- \
                                    --caliper-networkconfig connectors/(full|lite)/worker2.yaml \
                                    --caliper-benchconfig <benchmark config file>
(worker3.caliper.example.com) $ npm run worker -- \
                                    --caliper-networkconfig connectors/(full|lite)/worker3.yaml \
                                    --caliper-benchconfig <benchmark config file>
(worker4.caliper.example.com) $ npm run worker -- \
                                    --caliper-networkconfig connectors/(full|lite)/worker4.yaml \
                                    --caliper-benchconfig <benchmark config file>
(worker5.caliper.example.com) $ npm run worker -- \
                                    --caliper-networkconfig connectors/(full|lite)/worker5.yaml \
                                    --caliper-benchconfig <benchmark config file>

    The <benchmark config file> has to be same as the value used in the previous step. Also, the the parent directory name (full or lite) in connector configuration file must be the same as the parent directory name of the benchmark configuration file. For example, if you choose to run lite/service-broker/request.yaml, the connector configuration must be connectors/lite/worker*.yaml.