infra/proxyd/README.md

147 lines
5.3 KiB
Markdown
Raw Normal View History

# rpc-proxy
This tool implements `proxyd`, an RPC request router and proxy. It does the following things:
1. Whitelists RPC methods.
2. Routes RPC methods to groups of backend services.
3. Automatically retries failed backend requests.
2023-05-30 18:14:52 +03:00
4. Track backend consensus (`latest`, `safe`, `finalized` blocks), peer count and sync state.
5. Re-write requests and responses to enforce consensus.
6. Load balance requests across backend services.
7. Cache immutable responses from backends.
2023-12-13 20:50:18 +03:00
8. Provides metrics to measure request latency, error rates, and the like.
2023-05-30 18:14:52 +03:00
## Usage
Run `make proxyd` to build the binary. No additional dependencies are necessary.
To configure `proxyd` for use, you'll need to create a configuration file to define your proxy backends and routing rules. Check out [example.config.toml](./example.config.toml) for how to do this alongside a full list of all options with commentary.
Once you have a config file, start the daemon via `proxyd <path-to-config>.toml`.
2023-05-30 18:14:52 +03:00
## Consensus awareness
Starting on v4.0.0, `proxyd` is aware of the consensus state of its backends. This helps minimize chain reorgs experienced by clients.
To enable this behavior, you must set `consensus_aware` value to `true` in the backend group.
When consensus awareness is enabled, `proxyd` will poll the backends for their states and resolve a consensus group based on:
* the common ancestor `latest` block, i.e. if a backend is experiencing a fork, the fork won't be visible to the clients
* the lowest `safe` block
* the lowest `finalized` block
* peer count
* sync state
The backend group then acts as a round-robin load balancer distributing traffic equally across healthy backends in the consensus group, increasing the availability of the proxy.
A backend is considered healthy if it meets the following criteria:
* not banned
* avg 1-min moving window error rate ≤ configurable threshold
* avg 1-min moving window latency ≤ configurable threshold
* peer count ≥ configurable threshold
* `latest` block lag ≤ configurable threshold
* last state update ≤ configurable threshold
* not currently syncing
When a backend is experiencing inconsistent consensus, high error rates or high latency,
the backend will be banned for a configurable amount of time (default 5 minutes)
and won't receive any traffic during this period.
## Tag rewrite
When consensus awareness is enabled, `proxyd` will enforce the consensus state transparently for all the clients.
For example, if a client requests the `eth_getBlockByNumber` method with the `latest` tag,
`proxyd` will rewrite the request to use the resolved latest block from the consensus group
and forward it to the backend.
The following request methods are rewritten:
* `eth_getLogs`
* `eth_newFilter`
* `eth_getBalance`
* `eth_getCode`
* `eth_getTransactionCount`
* `eth_call`
* `eth_getStorageAt`
* `eth_getBlockTransactionCountByNumber`
* `eth_getUncleCountByBlockNumber`
* `eth_getBlockByNumber`
* `eth_getTransactionByBlockNumberAndIndex`
* `eth_getUncleByBlockNumberAndIndex`
* `debug_getRawReceipts`
2023-05-30 18:14:52 +03:00
And `eth_blockNumber` response is overridden with current block consensus.
## Cacheable methods
Cache use Redis and can be enabled for the following immutable methods:
* `eth_chainId`
* `net_version`
* `eth_getBlockTransactionCountByHash`
* `eth_getUncleCountByBlockHash`
* `eth_getBlockByHash`
* `eth_getTransactionByBlockHashAndIndex`
* `eth_getUncleByBlockHashAndIndex`
* `debug_getRawReceipts` (block hash only)
2023-05-30 18:14:52 +03:00
## Meta method `consensus_getReceipts`
To support backends with different specifications in the same backend group,
proxyd exposes a convenient method to fetch receipts abstracting away
what specific backend will serve the request.
Each backend specifies their preferred method to fetch receipts with `consensus_receipts_target` config,
which will be translated from `consensus_getReceipts`.
This method takes a `blockNumberOrHash` (i.e. `tag|qty|hash`)
and returns the receipts for all transactions in the block.
Request example
```json
{
"jsonrpc":"2.0",
"id": 1,
"params": ["0xc6ef2fc5426d6ad6fd9e2a26abeab0aa2411b7ab17f30a99d3cb96aed1d1055b"]
}
```
It currently supports translation to the following targets:
* `debug_getRawReceipts(blockOrHash)` (default)
* `alchemy_getTransactionReceipts(blockOrHash)`
* `parity_getBlockReceipts(blockOrHash)`
* `eth_getBlockReceipts(blockOrHash)`
The selected target is returned in the response, in a wrapped result.
Response example
```json
{
"jsonrpc": "2.0",
"id": 1,
"result": {
"method": "debug_getRawReceipts",
"result": {
// the actual raw result from backend
}
}
}
```
See [op-node receipt fetcher](https://github.com/ethereum-optimism/optimism/blob/186e46a47647a51a658e699e9ff047d39444c2de/op-node/sources/receipts.go#L186-L253).
## Metrics
2023-05-30 18:14:52 +03:00
See `metrics.go` for a list of all available metrics.
The metrics port is configurable via the `metrics.port` and `metrics.host` keys in the config.
## Adding Backend SSL Certificates in Docker
2023-05-30 18:14:52 +03:00
The Docker image runs on Alpine Linux. If you get SSL errors when connecting to a backend within Docker, you may need to add additional certificates to Alpine's certificate store. To do this, bind mount the certificate bundle into a file in `/usr/local/share/ca-certificates`. The `entrypoint.sh` script will then update the store with whatever is in the `ca-certificates` directory prior to starting `proxyd`.