client/network: Use request response for block requests (#7478)

* client/network: Add scaffolding for finality req to use req resp
	#sc

* client/network/src/finality_requests: Remove

* client/network/src/behaviour: Pass request id down to sync

* client/network: Use request response for block requests

* client/network: Move handler logic into *_*_handler.rs

* client/network: Track ongoing finality requests in protocol.rs

* client/network: Remove commented out finalization initialization

* client/network: Add docs for request handlers

* client/network/finality_request_handler: Log errors

* client/network/block_request_handler: Log errors

* client/network: Format

* client/network: Handle block request failure

* protocols/network: Fix tests

* client/network/src/behaviour: Handle request sending errors

* client/network: Move response handling into custom method

* client/network/protocol: Handle block response errors

* client/network/protocol: Remove tracking of obsolete requests

* client/network/protocol: Remove block request start time tracking

This will be handled generically via request-responses.

* client/network/protocol: Refactor on_*_request_started

* client/network: Pass protocol config instead of protocol name

* client/network: Pass protocol config in tests

* client/network/config: Document request response configs

* client/network/src/_request_handler: Document protocol config gen

* client/network/src/protocol: Document Peer request values

* client/network: Rework request response to always use oneshot

* client/network: Unified metric reporting for all request protocols

* client/network: Move protobuf parsing into protocol.rs

* client/network/src/protocol: Return pending events after poll

* client/network: Improve error handling and documentation

* client/network/behaviour: Remove outdated error types

* Update client/network/src/block_request_handler.rs

Co-authored-by: Ashley <ashley.ruglys@gmail.com>

* Update client/network/src/finality_request_handler.rs

Co-authored-by: Ashley <ashley.ruglys@gmail.com>

* client/network/protocol: Reduce reputation on timeout

* client/network/protocol: Refine reputation changes

* client/network/block_request_handler: Set and explain queue length

* client/service: Deny block requests when light client

* client/service: Fix role matching

* client: Enforce line width

* client/network/request_responses: Fix unit tests

* client/network: Expose time to build response via metrics

* client/network/request_responses: Fix early connection closed error

* client/network/protocol: Fix line length

* client/network/protocol: Disconnect on most request failures

* client/network/protocol: Disconnect peer when oneshot is canceled

* client/network/protocol: Disconnect peer even when connection closed

* client/network/protocol: Remove debugging log line

* client/network/request_response: Use Clone::clone for error

* client/network/request_response: Remove outdated comment

With libp2p v0.33.0 libp2p-request-response properly sends inbound
failures on connections being closed.

Co-authored-by: Addie Wagenknecht <addie@nortd.com>
Co-authored-by: Ashley <ashley.ruglys@gmail.com>
This commit is contained in:
Max Inden
2021-01-05 19:20:54 +01:00
committed by GitHub
parent 92f596829d
commit 3f629f743b
17 changed files with 780 additions and 1299 deletions
@@ -0,0 +1,220 @@
// Copyright 2020 Parity Technologies (UK) Ltd.
// This file is part of Substrate.
// Substrate is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
// Substrate is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
// You should have received a copy of the GNU General Public License
// along with Substrate. If not, see <http://www.gnu.org/licenses/>.
//! Helper for handling (i.e. answering) block requests from a remote peer via the
//! [`crate::request_responses::RequestResponsesBehaviour`].
use codec::{Encode, Decode};
use crate::chain::Client;
use crate::config::ProtocolId;
use crate::protocol::{message::BlockAttributes};
use crate::request_responses::{IncomingRequest, ProtocolConfig};
use crate::schema::v1::block_request::FromBlock;
use crate::schema::v1::{BlockResponse, Direction};
use futures::channel::{mpsc, oneshot};
use futures::stream::StreamExt;
use log::debug;
use prost::Message;
use sp_runtime::generic::BlockId;
use sp_runtime::traits::{Block as BlockT, Header, One, Zero};
use std::cmp::min;
use std::sync::{Arc};
use std::time::Duration;
const LOG_TARGET: &str = "block-request-handler";
const MAX_BLOCKS_IN_RESPONSE: usize = 128;
const MAX_BODY_BYTES: usize = 8 * 1024 * 1024;
/// Generates a [`ProtocolConfig`] for the block request protocol, refusing incoming requests.
pub fn generate_protocol_config(protocol_id: ProtocolId) -> ProtocolConfig {
ProtocolConfig {
name: generate_protocol_name(protocol_id).into(),
max_request_size: 1024 * 1024,
max_response_size: 16 * 1024 * 1024,
request_timeout: Duration::from_secs(40),
inbound_queue: None,
}
}
/// Generate the block protocol name from chain specific protocol identifier.
fn generate_protocol_name(protocol_id: ProtocolId) -> String {
let mut s = String::new();
s.push_str("/");
s.push_str(protocol_id.as_ref());
s.push_str("/sync/2");
s
}
/// Handler for incoming block requests from a remote peer.
pub struct BlockRequestHandler<B> {
client: Arc<dyn Client<B>>,
request_receiver: mpsc::Receiver<IncomingRequest>,
}
impl <B: BlockT> BlockRequestHandler<B> {
/// Create a new [`BlockRequestHandler`].
pub fn new(protocol_id: ProtocolId, client: Arc<dyn Client<B>>) -> (Self, ProtocolConfig) {
// Rate of arrival multiplied with the waiting time in the queue equals the queue length.
//
// An average Polkadot sentry node serves less than 5 requests per second. The 95th percentile
// serving a request is less than 2 second. Thus one would estimate the queue length to be
// below 10.
//
// Choosing 20 as the queue length to give some additional buffer.
let (tx, request_receiver) = mpsc::channel(20);
let mut protocol_config = generate_protocol_config(protocol_id);
protocol_config.inbound_queue = Some(tx);
(Self { client, request_receiver }, protocol_config)
}
fn handle_request(
&self,
payload: Vec<u8>,
pending_response: oneshot::Sender<Vec<u8>>
) -> Result<(), HandleRequestError> {
let request = crate::schema::v1::BlockRequest::decode(&payload[..])?;
let from_block_id = match request.from_block.ok_or(HandleRequestError::MissingFromField)? {
FromBlock::Hash(ref h) => {
let h = Decode::decode(&mut h.as_ref())?;
BlockId::<B>::Hash(h)
}
FromBlock::Number(ref n) => {
let n = Decode::decode(&mut n.as_ref())?;
BlockId::<B>::Number(n)
}
};
let max_blocks = if request.max_blocks == 0 {
MAX_BLOCKS_IN_RESPONSE
} else {
min(request.max_blocks as usize, MAX_BLOCKS_IN_RESPONSE)
};
let direction = Direction::from_i32(request.direction)
.ok_or(HandleRequestError::ParseDirection)?;
let attributes = BlockAttributes::from_be_u32(request.fields)?;
let get_header = attributes.contains(BlockAttributes::HEADER);
let get_body = attributes.contains(BlockAttributes::BODY);
let get_justification = attributes.contains(BlockAttributes::JUSTIFICATION);
let mut blocks = Vec::new();
let mut block_id = from_block_id;
let mut total_size: usize = 0;
while let Some(header) = self.client.header(block_id).unwrap_or(None) {
let number = *header.number();
let hash = header.hash();
let parent_hash = *header.parent_hash();
let justification = if get_justification {
self.client.justification(&BlockId::Hash(hash))?
} else {
None
};
let is_empty_justification = justification.as_ref().map(|j| j.is_empty()).unwrap_or(false);
let body = if get_body {
match self.client.block_body(&BlockId::Hash(hash))? {
Some(mut extrinsics) => extrinsics.iter_mut()
.map(|extrinsic| extrinsic.encode())
.collect(),
None => {
log::trace!(target: "sync", "Missing data for block request.");
break;
}
}
} else {
Vec::new()
};
let block_data = crate::schema::v1::BlockData {
hash: hash.encode(),
header: if get_header {
header.encode()
} else {
Vec::new()
},
body,
receipt: Vec::new(),
message_queue: Vec::new(),
justification: justification.unwrap_or_default(),
is_empty_justification,
};
total_size += block_data.body.len();
blocks.push(block_data);
if blocks.len() >= max_blocks as usize || total_size > MAX_BODY_BYTES {
break
}
match direction {
Direction::Ascending => {
block_id = BlockId::Number(number + One::one())
}
Direction::Descending => {
if number.is_zero() {
break
}
block_id = BlockId::Hash(parent_hash)
}
}
}
let res = BlockResponse { blocks };
let mut data = Vec::with_capacity(res.encoded_len());
res.encode(&mut data)?;
pending_response.send(data)
.map_err(|_| HandleRequestError::SendResponse)
}
/// Run [`BlockRequestHandler`].
pub async fn run(mut self) {
while let Some(request) = self.request_receiver.next().await {
let IncomingRequest { peer, payload, pending_response } = request;
match self.handle_request(payload, pending_response) {
Ok(()) => debug!(target: LOG_TARGET, "Handled block request from {}.", peer),
Err(e) => debug!(
target: LOG_TARGET,
"Failed to handle block request from {}: {}",
peer, e,
),
}
}
}
}
#[derive(derive_more::Display, derive_more::From)]
enum HandleRequestError {
#[display(fmt = "Failed to decode request: {}.", _0)]
DecodeProto(prost::DecodeError),
#[display(fmt = "Failed to encode response: {}.", _0)]
EncodeProto(prost::EncodeError),
#[display(fmt = "Failed to decode block hash: {}.", _0)]
DecodeScale(codec::Error),
#[display(fmt = "Missing `BlockRequest::from_block` field.")]
MissingFromField,
#[display(fmt = "Failed to parse BlockRequest::direction.")]
ParseDirection,
Client(sp_blockchain::Error),
#[display(fmt = "Failed to send response.")]
SendResponse,
}