Lavanet challenge

Lavanet challenge is an interview task for lava network. The challenge is structured in 2 parts

• Server - needs to support a forward proxy service which needs to be able to register gRPC service, forward the request to a specific network, and return the response
• Tracker - needs to connect to the server, listen for new events and prints test results

Architecture

Server

The server is the main service of the lavanet_challenge. It can be extended to support many services. This version supports only forward_proxy service

• Forward proxy needs to listen for gRPC requests, forwards them to a specific client, and return a response. This version of code support only cosmos.base.tendermint.v1beta1.Service but it is envisioned to be easily extended. If you wish to contribute and extend services here are the changes needed to be implemented:
  • Create a server and client implementation from desired service example
  • Extended forward proxy object to keep track of new client here
  • Register server here

To see all supported server commands:

go run main.go server --help

The default command that starts lavanet_challenge client

Usage:
   server [flags]

Flags:
      --grpc-address string           the GRPC interface (default "127.0.0.1:9632")
  -h, --help                          help for server
      --log-level string              the log level for console output (default "INFO")
      --log-to string                 write all logs to the file at specified location instead of writing them to console
      --network-grpc-address string   network gRPC endpoint (default "grpc.osmosis.zone:9090")

Tracker

Tracker is designed to support multiple specific trackers. This version supports only block tracker service.

• Block tracker is a scheduled polling machine whose goal is to print test results for a specified block range. It can be connected to any network which supports cosmos.base.tendermint.v1beta1.Service.GetLatestBlock api. Depending on the polling time 3 situations can occur.
  • New block arrived, it just needs to save it in the pending queue
  • The same block arrived, it will be skipped
  • Gap occurred so we need to fetch all blocks in between and save them i the pending queue

Once a pending queue gets full, the tests result will be exported in the format fileName_from_to.extension

To see all supported tracker commands:

go run main.go tracker --help

The default command that starts the tracker client

Usage:
   tracker [flags]

Flags:
      --client-grpc-address string   client gRPC endpoint (default "127.0.0.1:9632")
  -h, --help                         help for tracker
      --log-level string             the log level for console output (default "INFO")
      --log-to string                write all logs to the file at specified location instead of writing them to console
      --output-after uint            number after which results will be generated (default 5)
      --output-to string             write tracker data to the file at specified location (default "test_results.txt")
      --polling-time uint            polling time in seconds (default 2)

Run Locally

Clone the project

  git clone https://github.com/Aleksao998/lavanet_challenge.git

Go to the project directory

  cd lavanet_challenge

Start the server

  go run main.go server

** 2nd October osmosis mainnet stopped on 6263920, you can also use testnet **

  go run main.go server --network-grpc-address grpc-test.osmosis.zone:9090

Start the tracker

  go run main.go tracker

** To print test results for every new block use: **

  go run main.go tracker --output-after 1

To run linter

  make lint

To build binary

  make build

Manual testing

To test manually first thing you need to do is to install https://github.com/fullstorydev/grpcurl.

Steps to do:

• Start server with default flags (default server grpc address: localhost::9632)

  go run main.go server

• Start tracker (enable test results after every new block so you don't need to wait for 5 blocks)

  go run main.go tracker --output-after 1

To query the server you can use grpcurl commands:

Method	request
GetLatestBlock	grpcurl --plaintext localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetLatestBlock
GetLatestValidatorSet	grpcurl --plaintext localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetLatestValidatorSet
GetNodeInfo	grpcurl --plaintext localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetNodeInfo
GetSyncing	grpcurl --plaintext localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetSyncing
GetBlockByHeight	grpcurl --plaintext -d '{"height":5350708}' localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetBlockByHeight
GetValidatorSetByHeight	grpcurl --plaintext -d '{"height":5350708}' localhost:9632 cosmos.base.tendermint.v1beta1.Service.GetValidatorSetByHeight

Running tests

Lavanet_challenge has both e2e and unit tests.

• For the e2e test there is a freamwork which starts how many server instances we want. All e2e tests should be placed in the e2e folder
• For unit tests, there is a mock client for existing gRPC services

How to run:

• Build binary

  make buid

• Run all tests

  go test ./...

Want to contribute

• The lavanet_challenge uses a relatively basic feature proposition template, that is concise and to the point.

• For branching this project uses branch_type/branch_name format

• Before merging the PR, make sure that both lint and test git action passed

• Existing branching strategy and examples used so far can be found inside merged PR-s