TL;DR: A vote for Insight is about more than selecting a P-Rep node operator; it is a vote to leverage our powerful network of engineers and scientists to create secure infrastructure, contribute open-source developer tools, expand the universe of ICON dApps, and drive adoption at all levels of the ecosystem.
Who is Insight?
Insight is an international institute for higher education, offering funded Fellowships to help experienced software developers transition into specialized technical fields, such as blockchain engineering, DevOps, security, and data engineering. Each year, over 700 Insight Fellows build open source projects to demonstrate their skills and transition careers with leading tech firms.
Who is the team?
Mitchell Krawiec-Thayer, Ph.D. Decentralized Consensus Lead at Insight, Monero Research Lab
Rob Cannon DevOps engineering Fellow & Independent consultant
John Taylor DevOps Program Director at Insight
Jake Klamka Founder of Insight Data Science
Why vote for Insight?
Infrastructure: We have invested heavily in developing ICON’s infrastructure, and have had a full-time DevOps engineer working throughout Q3 2019 to deploy ICON infrastructure on the testnet, using DevOps best practices such as automated deployment and safe node architecture. All of our ICON tools are open source and free for use and modification. Other major projects include one-click ICON node deployment and the development of advanced logging infrastructure to enable developers and operators to quickly analyze and diagnose problems remotely without requiring/soliciting Telegram log dumps. This will particularly be useful when implementing new features and simulating the network under load.
dApps: Insight Fellowships consist of top engineers building useful projects, so we will attract new developers and onboard them directly into creating ICON integrations and decentralized applications. The high-quality B2B and B2C tools produced by Insight will drive ICON adoption by both businesses and end users - consequently increasing token value, to the benefit of both retail and institutional investors.
Documentation: We are generating technical documentation and tutorials geared toward ICON dApp developers and P-Rep node operators. Articles will be available in both English and Korean, to share best practices and help other node operators create secure implementations.
Given our infrastructure capability, educational contributions to ecosystem expansion, and continuous injection of dApp developers, it is a natural step to host one of the P-Rep nodes. We are excited about managing, governing, protecting, and growing the best network for blockchain interoperability.
Deploying P-Rep infrastructure as code (IaC) allows node maintainers to deploy their infrastructure repeatedly and consistently with minimal human intervention. We are evolving beyond boilerplate deployment steps into implementing and maintaining advanced security features as a community. Node maintainers and developers will also benefit from our advanced logging infrastructure that can be used to not only diagnose problems as they occur, but also give observability when conducting various stress and load tests in simulated environments.
Insight will build an open source modular reference architecture with Terraform, the leading IaC tool, wrapped with Terragrunt, in coordination with other major blockchain communities (waiting on acceptance of PR before announcements) and the developers of Terragrunt itself. The logging stack will be built on top of Terraform through a combination of Fluentd, Elasticsearch and Kibana with monitoring through Prometheus, InfluxDB, and Grafana. Deployment options ranging from a deployment through the docker-compose used to deploy the P-Rep node to a full Kubernetes cluster are being considered.
P-Rep Modular Reference Architecture
- Infrastructure as code modules to deploy single and clustered environments on Amazon Web Services
- Basics complete. Next steps are logging and security automation.
- Code generator to assemble modular reference architectures
- POC complete. Next step is rebuilding into a contained CLI. We are strategizing in collaboration with Gruntworks.
P-Rep Log Collection Infrastructure
- Immutable deployment of log collection and analysis infrastructure
- 3 months for basic stack
- Dashboarding system and log querying workflow to give operators observability over all key metrics
- 6 months
- Alerting API hooks and disaster recovery workflow and documentation
- 6 months
- Archival nodes and statistical analysis of network phenomena, both on-chain & metadata
- 6 months post-mainnet launch (need data to enable the research)
Wire or Telegram: @InsightMitchell