The Glassnode Code Challenge is an opportunity to demonstrate your problem solving skills which we value a lot at Glassnode.

The challenge is specified a bit loosely on purpose to let you demonstrate your ability to take intermediate decisions by your own and your ability to explore a previously unknown topic by yourself. Results of this challenge will be an entrypoint for the further tech interview.

Users of the Ethereum network are required to pay for each action a certain amount of ETH (called fees) roughly according to the following formula: kind of action * the gas price. Generally, the gas price increases in times of high activity/overloading, consequently increasing the total amount of fees spent by users. Ethereum recently experienced a significant increase in gas prices. It's interesting to analyze how much fees were spent by users on different kinds of actions. This can give insight into what kinds of actions were responsible for the increased gas prices. We can do this, by first filtering the transactions on the kind we are interested in and then summing the fees spent on those transactions.

For this task, we're interested in how much fees in the Ethereum network have been spent by plain ETH transfers. So we want you to compute the hourly amount of fees spent by transactions between externaly owned accounts (EOA). A transaction is considered to be between two EOA addresses if it's a direct ETH value transfer, i.e. to and from addresses of such a transaction should not be one of the contracts and not a special address 0x0000000000000000000000000000000000000000 used for contract creation. Fee computation is done in the following fashion: gas_used * gas_price. And of course we need some API to serve that information to the public.

Provided repository includes a docker-compose.yaml (docker-compose docs) with a database service in it, which represents a preconfigured postgres database with a data snapshot of Ethereum transactions for a single day (07.09.2020). Contract addresses are present in the provided database in the contracts table, and transactions themselves in the transactions table. Note, that gas_price is stored in Wei units in the data snapshot.

The end solution should

• include a REST API service listening on the port 8080 with an endpoint which serves data in the following JSON format:

[
  {
    "t": 1603114500,
    "v": 123.45
  },
  ...
]

where t is a unix timestamp of the hour, and v is the amount of fees being paid for transactions between EOA addresses within that hour in ETH units.

• be added to the services list in the provided docker-compose.yaml and the solution must be able to start using docker-compose up command

• be production ready

You're free to choose the path and parameters for the endpoint as well as the implementation language (but prefereably it should be golang).

• We respect your time and the challenge is designed in such a way as not to take more than 3-4 hours.
• In your repo create Solution.md with a description of your reasoning behind technical choices: trade-offs you might have made, anything you left out, or what you might do differently if you would had additional time.