Add request and response ID to every p2p message and route messages a… (

kaspanet#311) * increase initial difficulty 100x * Add request and response ID to every p2p message and route messages accordingly * Add protocol v6 * Some fixes * Address review comments * Add BLANK_ROUTE_ID --------- Co-authored-by: Michael Sutton <[email protected]>
aspectron · Nov 2, 2023 · 5556789 · 5556789
1 parent d66bd6c
commit 5556789
Show file tree

Hide file tree

Showing 18 changed files with 388 additions and 86 deletions.
diff --git a/consensus/core/src/config/genesis.rs b/consensus/core/src/config/genesis.rs
@@ -143,10 +143,10 @@ pub const TESTNET_GENESIS: GenesisBlock = GenesisBlock {
 
 pub const TESTNET11_GENESIS: GenesisBlock = GenesisBlock {
     hash: Hash::from_bytes([
-        0xf0, 0xa3, 0x15, 0x04, 0xfb, 0x4b, 0xf8, 0x83, 0xd0, 0xdd, 0x2a, 0x75, 0x1c, 0x7d, 0xac, 0x06, 0x1a, 0xb4, 0xc6, 0x33, 0x00,
-        0xc5, 0x4f, 0xd2, 0x64, 0xd8, 0xe3, 0x06, 0x99, 0xda, 0x51, 0x54,
+        0x5a, 0x90, 0xf8, 0x71, 0x09, 0x32, 0x3d, 0x61, 0x41, 0xff, 0x51, 0x04, 0xa2, 0xd5, 0xf8, 0xd8, 0x85, 0x7a, 0x6f, 0x39, 0x2e,
+        0xb4, 0x90, 0x5c, 0xe3, 0x55, 0x5e, 0xc9, 0x12, 0xcd, 0xfb, 0x9c,
     ]),
-    bits: 520421375, // see `gen_testnet11_genesis`
+    bits: 504155340, // see `gen_testnet11_genesis`
     ..TESTNET_GENESIS
 };
 
@@ -227,11 +227,11 @@ mod tests {
         let bps = Testnet11Bps::bps();
         let mut genesis = TESTNET_GENESIS;
         let target = kaspa_math::Uint256::from_compact_target_bits(genesis.bits);
-        let scaled_up_target = target * bps;
-        let scaled_up_bits = scaled_up_target.compact_target_bits();
-        genesis.bits = scaled_up_bits;
+        let scaled_target = target * bps / 100;
+        let scaled_bits = scaled_target.compact_target_bits();
+        genesis.bits = scaled_bits;
         if genesis.bits != TESTNET11_GENESIS.bits {
-            panic!("Testnet 11: new bits: {}\nnew hash: {:#04x?}", scaled_up_bits, Block::from(&genesis).hash().as_bytes());
+            panic!("Testnet 11: new bits: {}\nnew hash: {:#04x?}", scaled_bits, Block::from(&genesis).hash().as_bytes());
         }
     }
 

diff --git a/protocol/flows/src/flow_context.rs b/protocol/flows/src/flow_context.rs
@@ -1,5 +1,5 @@
 use crate::flowcontext::{orphans::OrphanBlocksPool, process_queue::ProcessQueue, transactions::TransactionsSpread};
-use crate::v5;
+use crate::{v5, v6};
 use async_trait::async_trait;
 use futures::future::join_all;
 use kaspa_addressmanager::AddressManager;
@@ -54,7 +54,7 @@ use tokio_stream::{wrappers::UnboundedReceiverStream, StreamExt};
 use uuid::Uuid;
 
 /// The P2P protocol version. Currently the only one supported.
-const PROTOCOL_VERSION: u32 = 5;
+const PROTOCOL_VERSION: u32 = 6;
 
 /// See `check_orphan_resolution_range`
 const BASELINE_ORPHAN_RESOLUTION_RANGE: u32 = 5;
@@ -573,8 +573,8 @@ impl ConnectionInitializer for FlowContext {
 
         // Register all flows according to version
         let (flows, applied_protocol_version) = match peer_version.protocol_version {
-            PROTOCOL_VERSION => (v5::register(self.clone(), router.clone()), PROTOCOL_VERSION),
-            // TODO: different errors for obsolete (low version) vs unknown (high)
+            v if v >= PROTOCOL_VERSION => (v6::register(self.clone(), router.clone()), PROTOCOL_VERSION),
+            5 => (v5::register(self.clone(), router.clone()), 5),
             v => return Err(ProtocolError::VersionMismatch(PROTOCOL_VERSION, v)),
         };
 
@@ -592,7 +592,7 @@ impl ConnectionInitializer for FlowContext {
         // Send and receive the ready signal
         handshake.exchange_ready_messages().await?;
 
-        info!("Registering p2p flows for peer {} for protocol version {}", router, peer_version.protocol_version);
+        info!("Registering p2p flows for peer {} for protocol version {}", router, applied_protocol_version);
 
         // Launch all flows. Note we launch only after the ready signal was exchanged
         for flow in flows {

diff --git a/protocol/flows/src/lib.rs b/protocol/flows/src/lib.rs
@@ -3,3 +3,4 @@ pub mod flow_trait;
 pub mod flowcontext;
 pub mod service;
 pub mod v5;
+pub mod v6;
diff --git a/protocol/flows/src/v5/blockrelay/flow.rs b/protocol/flows/src/v5/blockrelay/flow.rs
@@ -8,9 +8,9 @@ use kaspa_core::{debug, info};
 use kaspa_hashes::Hash;
 use kaspa_p2p_lib::{
     common::ProtocolError,
-    dequeue, dequeue_with_timeout, make_message,
+    dequeue, dequeue_with_timeout, make_message, make_request,
     pb::{kaspad_message::Payload, InvRelayBlockMessage, RequestBlockLocatorMessage, RequestRelayBlocksMessage},
-    IncomingRoute, Router,
+    IncomingRoute, Router, SharedIncomingRoute,
 };
 use std::{collections::VecDeque, sync::Arc};
 use tokio::sync::mpsc::{error::TrySendError, Sender};
@@ -22,12 +22,12 @@ pub struct RelayInvMessage {
 
 /// Encapsulates an incoming invs route which also receives data locally
 pub struct TwoWayIncomingRoute {
-    incoming_route: IncomingRoute,
+    incoming_route: SharedIncomingRoute,
     indirect_invs: VecDeque<Hash>,
 }
 
 impl TwoWayIncomingRoute {
-    pub fn new(incoming_route: IncomingRoute) -> Self {
+    pub fn new(incoming_route: SharedIncomingRoute) -> Self {
         Self { incoming_route, indirect_invs: VecDeque::new() }
     }
 
@@ -72,7 +72,7 @@ impl HandleRelayInvsFlow {
     pub fn new(
         ctx: FlowContext,
         router: Arc<Router>,
-        invs_route: IncomingRoute,
+        invs_route: SharedIncomingRoute,
         msg_route: IncomingRoute,
         ibd_sender: Sender<Block>,
     ) -> Self {
@@ -111,7 +111,7 @@ impl HandleRelayInvsFlow {
             }
 
             // We keep the request scope alive until consensus processes the block
-            let Some((block, request_scope)) = self.request_block(inv.hash).await? else {
+            let Some((block, request_scope)) = self.request_block(inv.hash, self.msg_route.id()).await? else {
                 debug!("Relay block {} was already requested from another peer, continuing...", inv.hash);
                 continue;
             };
@@ -184,13 +184,21 @@ impl HandleRelayInvsFlow {
         }
     }
 
-    async fn request_block(&mut self, requested_hash: Hash) -> Result<Option<(Block, RequestScope<Hash>)>, ProtocolError> {
+    async fn request_block(
+        &mut self,
+        requested_hash: Hash,
+        request_id: u32,
+    ) -> Result<Option<(Block, RequestScope<Hash>)>, ProtocolError> {
         // Note: the request scope is returned and should be captured until block processing is completed
         let Some(request_scope) = self.ctx.try_adding_block_request(requested_hash) else {
             return Ok(None);
         };
         self.router
-            .enqueue(make_message!(Payload::RequestRelayBlocks, RequestRelayBlocksMessage { hashes: vec![requested_hash.into()] }))
+            .enqueue(make_request!(
+                Payload::RequestRelayBlocks,
+                RequestRelayBlocksMessage { hashes: vec![requested_hash.into()] },
+                request_id
+            ))
             .await?;
         let msg = dequeue_with_timeout!(self.msg_route, Payload::Block)?;
         let block: Block = msg.try_into()?;
@@ -210,7 +218,7 @@ impl HandleRelayInvsFlow {
         // Add the block to the orphan pool if it's within orphan resolution range.
         // If the block is indirect it means one of its descendants was already is resolution range, so
         // we can avoid the query.
-        if is_indirect_inv || self.check_orphan_resolution_range(consensus, block.hash()).await? {
+        if is_indirect_inv || self.check_orphan_resolution_range(consensus, block.hash(), self.msg_route.id()).await? {
             let hash = block.hash();
             self.ctx.add_orphan(block).await;
             self.enqueue_orphan_roots(consensus, hash).await;
@@ -230,11 +238,17 @@ impl HandleRelayInvsFlow {
     /// mechanism or via IBD. This method sends a BlockLocator request to the peer with
     /// a limit of `ctx.orphan_resolution_range`. In the response, if we know none of the hashes,
     /// we should retrieve the given block `hash` via IBD. Otherwise, via unorphaning.
-    async fn check_orphan_resolution_range(&mut self, consensus: &ConsensusProxy, hash: Hash) -> Result<bool, ProtocolError> {
+    async fn check_orphan_resolution_range(
+        &mut self,
+        consensus: &ConsensusProxy,
+        hash: Hash,
+        request_id: u32,
+    ) -> Result<bool, ProtocolError> {
         self.router
-            .enqueue(make_message!(
+            .enqueue(make_request!(
                 Payload::RequestBlockLocator,
-                RequestBlockLocatorMessage { high_hash: Some(hash.into()), limit: self.ctx.orphan_resolution_range() }
+                RequestBlockLocatorMessage { high_hash: Some(hash.into()), limit: self.ctx.orphan_resolution_range() },
+                request_id
             ))
             .await?;
         let msg = dequeue_with_timeout!(self.msg_route, Payload::BlockLocator)?;

diff --git a/protocol/flows/src/v5/blockrelay/handle_requests.rs b/protocol/flows/src/v5/blockrelay/handle_requests.rs
@@ -2,7 +2,7 @@ use crate::{flow_context::FlowContext, flow_trait::Flow};
 use kaspa_core::debug;
 use kaspa_p2p_lib::{
     common::ProtocolError,
-    dequeue, make_message,
+    dequeue_with_request_id, make_message, make_response,
     pb::{kaspad_message::Payload, InvRelayBlockMessage},
     IncomingRoute, Router,
 };
@@ -36,14 +36,14 @@ impl HandleRelayBlockRequests {
         // Note: in go-kaspad this was done via a dedicated one-time flow.
         self.send_sink().await?;
         loop {
-            let msg = dequeue!(self.incoming_route, Payload::RequestRelayBlocks)?;
+            let (msg, request_id) = dequeue_with_request_id!(self.incoming_route, Payload::RequestRelayBlocks)?;
             let hashes: Vec<_> = msg.try_into()?;
 
             let session = self.ctx.consensus().unguarded_session();
 
             for hash in hashes {
                 let block = session.async_get_block(hash).await?;
-                self.router.enqueue(make_message!(Payload::Block, (&block).into())).await?;
+                self.router.enqueue(make_response!(Payload::Block, (&block).into(), request_id)).await?;
                 debug!("relayed block with hash {} to peer {}", hash, self.router);
             }
         }

diff --git a/protocol/flows/src/v5/mod.rs b/protocol/flows/src/v5/mod.rs
@@ -15,22 +15,22 @@ use self::{
 };
 use crate::{flow_context::FlowContext, flow_trait::Flow};
 
-use kaspa_p2p_lib::{KaspadMessagePayloadType, Router};
+use kaspa_p2p_lib::{KaspadMessagePayloadType, Router, SharedIncomingRoute};
 use std::sync::Arc;
 
-mod address;
-mod blockrelay;
-mod ibd;
-mod ping;
-mod request_anticone;
-mod request_block_locator;
-mod request_headers;
-mod request_ibd_blocks;
-mod request_ibd_chain_block_locator;
-mod request_pp_proof;
-mod request_pruning_point_and_anticone;
-mod request_pruning_point_utxo_set;
-mod txrelay;
+pub(crate) mod address;
+pub(crate) mod blockrelay;
+pub(crate) mod ibd;
+pub(crate) mod ping;
+pub(crate) mod request_anticone;
+pub(crate) mod request_block_locator;
+pub(crate) mod request_headers;
+pub(crate) mod request_ibd_blocks;
+pub(crate) mod request_ibd_chain_block_locator;
+pub(crate) mod request_pp_proof;
+pub(crate) mod request_pruning_point_and_anticone;
+pub(crate) mod request_pruning_point_utxo_set;
+pub(crate) mod txrelay;
 
 pub fn register(ctx: FlowContext, router: Arc<Router>) -> Vec<Box<dyn Flow>> {
     // IBD flow <-> invs flow channel requires no buffering hence the minimal size possible
@@ -60,7 +60,9 @@ pub fn register(ctx: FlowContext, router: Arc<Router>) -> Vec<Box<dyn Flow>> {
         Box::new(HandleRelayInvsFlow::new(
             ctx.clone(),
             router.clone(),
-            router.subscribe_with_capacity(vec![KaspadMessagePayloadType::InvRelayBlock], ctx.block_invs_channel_size()),
+            SharedIncomingRoute::new(
+                router.subscribe_with_capacity(vec![KaspadMessagePayloadType::InvRelayBlock], ctx.block_invs_channel_size()),
+            ),
             router.subscribe(vec![KaspadMessagePayloadType::Block, KaspadMessagePayloadType::BlockLocator]),
             ibd_sender,
         )),

diff --git a/protocol/flows/src/v5/request_anticone.rs b/protocol/flows/src/v5/request_anticone.rs
@@ -4,7 +4,7 @@ use kaspa_core::debug;
 use kaspa_hashes::Hash;
 use kaspa_p2p_lib::{
     common::ProtocolError,
-    dequeue, make_message,
+    dequeue_with_request_id, make_response,
     pb::{kaspad_message::Payload, BlockHeadersMessage, DoneHeadersMessage},
     IncomingRoute, Router,
 };
@@ -34,7 +34,7 @@ impl HandleAnticoneRequests {
 
     async fn start_impl(&mut self) -> Result<(), ProtocolError> {
         loop {
-            let msg = dequeue!(self.incoming_route, Payload::RequestAnticone)?;
+            let (msg, request_id) = dequeue_with_request_id!(self.incoming_route, Payload::RequestAnticone)?;
             let (block, context): (Hash, Hash) = msg.try_into()?;
 
             debug!("received anticone request with block hash: {}, context hash: {} for peer {}", block, context, self.router);
@@ -55,12 +55,13 @@ impl HandleAnticoneRequests {
             headers.sort_by(|a, b| a.blue_work.cmp(&b.blue_work));
 
             self.router
-                .enqueue(make_message!(
+                .enqueue(make_response!(
                     Payload::BlockHeaders,
-                    BlockHeadersMessage { block_headers: headers.into_iter().map(|header| header.as_ref().into()).collect() }
+                    BlockHeadersMessage { block_headers: headers.into_iter().map(|header| header.as_ref().into()).collect() },
+                    request_id
                 ))
                 .await?;
-            self.router.enqueue(make_message!(Payload::DoneHeaders, DoneHeadersMessage {})).await?;
+            self.router.enqueue(make_response!(Payload::DoneHeaders, DoneHeadersMessage {}, request_id)).await?;
         }
     }
 }
diff --git a/protocol/flows/src/v5/request_block_locator.rs b/protocol/flows/src/v5/request_block_locator.rs
@@ -2,7 +2,7 @@ use std::sync::Arc;
 
 use kaspa_p2p_lib::{
     common::ProtocolError,
-    dequeue, make_message,
+    dequeue_with_request_id, make_response,
     pb::{kaspad_message::Payload, BlockLocatorMessage},
     IncomingRoute, Router,
 };
@@ -33,16 +33,17 @@ impl RequestBlockLocatorFlow {
 
     async fn start_impl(&mut self) -> Result<(), ProtocolError> {
         loop {
-            let msg = dequeue!(self.incoming_route, Payload::RequestBlockLocator)?;
+            let (msg, request_id) = dequeue_with_request_id!(self.incoming_route, Payload::RequestBlockLocator)?;
             let (high, limit) = msg.try_into()?;
 
             let locator =
                 self.ctx.consensus().session().await.async_create_block_locator_from_pruning_point(high, limit as usize).await?;
 
             self.router
-                .enqueue(make_message!(
+                .enqueue(make_response!(
                     Payload::BlockLocator,
-                    BlockLocatorMessage { hashes: locator.into_iter().map(|hash| hash.into()).collect() }
+                    BlockLocatorMessage { hashes: locator.into_iter().map(|hash| hash.into()).collect() },
+                    request_id
                 ))
                 .await?;
         }

diff --git a/protocol/flows/src/v5/request_headers.rs b/protocol/flows/src/v5/request_headers.rs
@@ -4,7 +4,7 @@ use kaspa_consensus_core::api::ConsensusApi;
 use kaspa_hashes::Hash;
 use kaspa_p2p_lib::{
     common::ProtocolError,
-    dequeue, make_message,
+    dequeue, dequeue_with_request_id, make_response,
     pb::{self, kaspad_message::Payload, BlockHeadersMessage, DoneHeadersMessage},
     IncomingRoute, Router,
 };
@@ -40,7 +40,7 @@ impl RequestHeadersFlow {
         let max_blocks = max(MAX_BLOCKS, self.ctx.config.mergeset_size_limit as usize + 1);
 
         loop {
-            let msg = dequeue!(self.incoming_route, Payload::RequestHeaders)?;
+            let (msg, request_id) = dequeue_with_request_id!(self.incoming_route, Payload::RequestHeaders)?;
             let (high, mut low) = msg.try_into()?;
 
             let consensus = self.ctx.consensus();
@@ -68,13 +68,13 @@ impl RequestHeadersFlow {
                     session.spawn_blocking(move |c| Self::get_headers_between(c, low, high, max_blocks)).await?;
                 debug!("Got {} header hashes above {}", block_headers.len(), low);
                 low = last;
-                self.router.enqueue(make_message!(Payload::BlockHeaders, BlockHeadersMessage { block_headers })).await?;
+                self.router.enqueue(make_response!(Payload::BlockHeaders, BlockHeadersMessage { block_headers }, request_id)).await?;
 
                 dequeue!(self.incoming_route, Payload::RequestNextHeaders)?;
                 session = consensus.session().await;
             }
 
-            self.router.enqueue(make_message!(Payload::DoneHeaders, DoneHeadersMessage {})).await?;
+            self.router.enqueue(make_response!(Payload::DoneHeaders, DoneHeadersMessage {}, request_id)).await?;
         }
     }
 

diff --git a/protocol/flows/src/v5/request_ibd_blocks.rs b/protocol/flows/src/v5/request_ibd_blocks.rs
@@ -1,6 +1,8 @@
 use crate::{flow_context::FlowContext, flow_trait::Flow};
 use kaspa_core::debug;
-use kaspa_p2p_lib::{common::ProtocolError, dequeue, make_message, pb::kaspad_message::Payload, IncomingRoute, Router};
+use kaspa_p2p_lib::{
+    common::ProtocolError, dequeue_with_request_id, make_response, pb::kaspad_message::Payload, IncomingRoute, Router,
+};
 use std::sync::Arc;
 
 pub struct HandleIbdBlockRequests {
@@ -27,15 +29,15 @@ impl HandleIbdBlockRequests {
 
     async fn start_impl(&mut self) -> Result<(), ProtocolError> {
         loop {
-            let msg = dequeue!(self.incoming_route, Payload::RequestIbdBlocks)?;
+            let (msg, request_id) = dequeue_with_request_id!(self.incoming_route, Payload::RequestIbdBlocks)?;
             let hashes: Vec<_> = msg.try_into()?;
 
             debug!("got request for {} IBD blocks", hashes.len());
             let session = self.ctx.consensus().unguarded_session();
 
             for hash in hashes {
                 let block = session.async_get_block(hash).await?;
-                self.router.enqueue(make_message!(Payload::IbdBlock, (&block).into())).await?;
+                self.router.enqueue(make_response!(Payload::IbdBlock, (&block).into(), request_id)).await?;
             }
         }
     }