ICON Blockchain Centralized Logging¶
THIS PROPOSAL IS A WORK IN PROGRESS
This might be too simple of a project for a fellow to take on.
This project aims to centralize the logging collection from the P-Rep, Citizen nodes, and sentry nodes for the ICON network.
Business case idea¶
We are currently developing numerous different configurations of P-Rep nodes that we will be testing under load to determine the reliability of the network. During these tests, we will be intentionally introducing numerous different kinds of stress to the network and expect all kinds of different failures to occur. The only way to diagnose the root cause of these failures is if we have a robust log collection infrastructure. This project will empower numerous different workstreams that depend on those logs, thereby increasing development productivity as we gain observability into the application layer of the network.
This project has two components.
Basic version running that is a minimal configuration to run in an issolated environment that tracks
More advanced configuration - tbd
Minimal Configuration Challenges¶
The simple solution could be that we just run a node and spin up this docker-compose in
Setting up agent config from the application nodes
We should start to nail down what streams are coming in and how to analyze them in Kibana
Spinning up terraform
Rob will help with this. You should have a copy of the stack you need when you start the project
Dealing with secrets appropriately
We’ll need to fiugre out a short and long term plan for secrets
Short term we’ll be going for the easiest approach and we’ll just tranfer keys manually
Long term we need a simple solution that the community feels comfortable with
Kubernetes Configuration Challenges¶
terraform + terragrunt (basic architecture supplied)
Simple setup with helm