Cow Protocol Services

This repository contains backend code for Cow Protocol Services written in Rust.

Order Book

The orderbook crate provides the http api through which users (usually through a frontend web application) interact with the order book. Users can add signed orders to the order book and query the state of their orders. They can also use the API to estimate fee amounts and limit prices before placing their order.

Solvers also interact with the order book by querying a list of open orders that they can attempt to settle.

The api is documented with openapi. A simple example script that uses the API to place random orders can be found in this repo

The order book service itself uses PostgreSQL as a backend to persist orders. In addition to connecting the http api to the database it also checks order validity based on the block time, trade events, erc20 funding and approval so that solvers can query only valid orders.

Multiple concurrent orderbooks can run at the same time, allowing the user-facing API to scale horizontally with increased traffic.

Autopilot

The autopilot crate is responsible for driving the protocol forward. Concretely, it is responsible for "cutting" new auctions (i.e. determining auction boundaries and which orders are to be included, as well as various parameters important for settlement objective value computation).

The autopilot connects to the same PostgreSQL database as the orderbook and uses it to query orders as well as storing the most recent auction and settlement competition.

Solver

The solver crate is responsible for submitting on-chain settlements based on the orders it gets from the order book and other liquidity sources like Balancer or Uniswap pools.

It implements a few settlement strategies directly in Rust:

• Naive Solver: Can match to overlapping opposing orders (e.g. DAI for WETH & WETH for DAI) with one another settling the excess with Uniswap
• Uniswap Baseline: Same path finding as used by the Uniswap frontend (settling orders individually instead of batching them together)

It can also interact with a more advanced, Gnosis internal, closed source solver which tries to settle all orders using the combinatorial optimization formulations described in Multi-Token Batch Auctions with Uniform Clearing Price

Other Crates

There are additional crates that live in the cargo workspace.

• alerter provides a custom alerter binary that looks at the current orderbook and counts metrics for orders that should be solved but aren't
• contract provides ethcontract-rs based smart contract bindings
• database provides the shared database and storage layer logic shared between the autopilot and orderbook
• driver an in-development binary that intends to replace the solver; it has a slightly different design that allows co-location with external solvers
• e2e end-to-end tests
• ethrpc ethrpc client with a few extensions
• model provides the serialization model for orders in the order book api
• number extensions to number types, such as numerical conversions between 256-bit integers, nonzero types and de/serialization implementations
• observe initialization and helper functions for logging and metrics
• shared provides other shared functionality between the solver and order book
• testlib shared helpers for writing unit and end-to-end tests

Testing

The CI (check .github/workflows/pull-request.yaml) runs unit tests, e2e tests, clippy and cargo fmt

Unit Tests:

cargo test

DB Tests:

cargo test -- postgres --test-threads 1 --ignored

Note: Requires postgres database running (see below).

E2E Tests - Local Node:

cargo test -p e2e local_node -- --ignored.

Note: Requires postgres database and local test network with smart contracts deployed (see below).

E2E Tests - Forked Node:

FORK_URL=<mainnet archive node RPC URL> cargo test -p e2e forked_node -- --ignored.

Note: Requires postgres database (see below).

Clippy

cargo clippy --all-features --all-targets -- -D warnings

Flaky Tests

In case a test is flaky and only fails sometimes in CI you can use the run-flaky-test github action to test your fix with the CI to get confidence that the fix that works locally also works in CI.

Development Setup

Postgres

The tests that require postgres connect to the default database of a locally running postgres instance on the default port. To achieve this, open a new shell and run the command below: Note: The migrations will be applied as well.

docker-compose up

Local Test Network

In order to run the e2e tests you have to have an EVM compatible testnet running locally. We make use of anvil from the Foundry project to spin up a local testnet.

anvil supports all the RPC methods we need to run the services and tests.

1. Install foundryup.
2. Install foundry with foundryup.
3. Run anvil with the following configuration:

ANVIL_IP_ADDR=0.0.0.0 anvil \
  --gas-price 1 \
  --gas-limit 10000000 \
  --base-fee 0 \
  --balance 1000000 \
  --chain-id 1 \
  --timestamp 1577836800

Profiling

All binaries are compiled with support for tokio-console by default to allow you to look inside the tokio runtime. However, this feature is not enabled at runtime by default because it comes with a pretty significant memory overhead. To enable it you just have to set the environment variable TOKIO_CONSOLE=true and run the binary you want to instrument.

You can install and run tokio-console with:

cargo install --locked tokio-console
tokio-console

Changing Log Filters

It's possible to change the tracing log filter while the process is running. This can be useful to debug an error that requires more verbose logs but which might no longer appear after restarting the system.

Each process opens a UNIX socket at /tmp/log_filter_override_<program_name>_<pid>.sock. To change the log filter connect to it with nc -U <path> and enter a new log filter. You can also reset the log filter to the filter the program was initially started with by entering reset.

See here for documentation on the supported log filter format.

Running the Services Locally

Prerequisites

Reading the state of the blockchain requires issuing RPC calls to an ethereum node. This can be a testnet you are running locally, some "real" node you have access to or the most convenient thing is to use a third-party service like infura to get access to an ethereum node which we recommend. After you made a free infura account they offer you "endpoints" for the mainnet and different testnets. We will refer those as node-urls. Because services are only run on Mainnet, Gnosis Chain, Arbitrum One and Sepolia you need to select one of those.

Note that the node-url is sensitive data. The orderbook and solver executables allow you to pass it with the --node-url parameter. This is very convenient for our examples but to minimize the possibility of sharing this information by accident you should consider setting the NODE_URL environment variable so you don't have to pass the --node-url argument to the executables.

To avoid confusion during your tests, always double-check that the token and account addresses you use actually correspond to the network of the node-url you are running the executables with.

Autopilot

To see all supported command line arguments run cargo run --bin autopilot -- --help.

Run an autopilot with:

cargo run --bin autopilot -- \
  --skip-event-sync true \
  --node-url <YOUR_NODE_URL>

--skip-event-sync will skip some work to speed up the initialization process.

Orderbook

To see all supported command line arguments run cargo run --bin orderbook -- --help.

Run an orderbook on localhost:8080 with:

cargo run --bin orderbook -- \
  --node-url <YOUR_NODE_URL>

If your node supports trace_callMany, or you have an additional node with tracing support, consider also specifying --tracing-node-url <YOUR_NODE_URL>. This will enable the tracing-based bad token detection.

Note: Current version of the code does not compile under Windows OS. Context and workaround are here.

Solvers

To see all supported command line arguments run cargo run --bin solver -- --help.

Run a solver which is connected to an orderbook at localhost:8080 with:

cargo run -p solver -- \
  --solver-account 0xa6DDBD0dE6B310819b49f680F65871beE85f517e \
  --transaction-strategy DryRun \
  --node-url <YOUR_NODE_URL>

--transaction-strategy DryRun will make the solver only print the solution but not submit it on-chain. This command is absolutely safe and will not use any funds.

The solver-account is responsible for signing transactions. Solutions for settlements need to come from an address the settlement contract trusts in order to make the contract actually consider the solution. If we pass a public address, like we do here, the solver only pretends to be used for testing purposes. To actually submit transactions on behalf of a solver account you would have to pass a private key of an account the settlement contract trusts instead. Adding your personal solver account is quite involved and requires you to get in touch with the team, so we are using this public solver address for now.

To make things more interesting and see some real orders you can connect the solver to our real orderbook service. There are several orderbooks for production and staging environments on different networks. Find the orderbook-url corresponding to your node-url which suits your purposes and connect your solver to it with --orderbook-url <URL>.

Always make sure that the solver and the orderbook it connects to are configured to use the same network.

Frontend

To conveniently submit orders checkout the CowSwap frontend and point it to your local instance.