first draft

Cargo.lock

@@ -550,11 +550,11 @@ version = "0.1.0"
 dependencies = [
  "borsh",
  "cuprate-constants",
+ "cuprate-p2p-bucket",
  "cuprate-p2p-core",
  "cuprate-pruning",
  "cuprate-test-utils",
  "futures",
- "indexmap",
  "rand",
  "thiserror",
  "tokio",
@@ -871,6 +871,7 @@ dependencies = [
  "borsh",
  "cfg-if",
  "cuprate-helper",
+ "cuprate-p2p-bucket",
  "cuprate-pruning",
  "cuprate-test-utils",
  "cuprate-types",
@@ -936,6 +937,7 @@ dependencies = [
  "async-trait",
  "borsh",
  "cuprate-helper",
+ "cuprate-p2p-bucket",
  "cuprate-p2p-core",
  "cuprate-types",
  "cuprate-wire",

Cargo.toml

@@ -65,7 +65,7 @@ cuprate-levin               = { path = "net/levin"              ,default-feature
 cuprate-wire                = { path = "net/wire"               ,default-features = false}
 cuprate-p2p                 = { path = "p2p/p2p"                ,default-features = false}
 cuprate-p2p-core            = { path = "p2p/p2p-core"           ,default-features = false}
-cuprate-p2p-bucket          = { path = "p2p/p2p-bucket"         ,default-features = false}
+cuprate-p2p-bucket          = { path = "p2p/bucket"             ,default-features = false}
 cuprate-dandelion-tower     = { path = "p2p/dandelion-tower"    ,default-features = false}
 cuprate-async-buffer        = { path = "p2p/async-buffer"       ,default-features = false}
 cuprate-address-book        = { path = "p2p/address-book"       ,default-features = false}

p2p/address-book/Cargo.toml

@@ -10,6 +10,7 @@ authors = ["Boog900"]
 cuprate-constants   = { workspace = true }
 cuprate-pruning     = { workspace = true }
 cuprate-p2p-core    = { workspace = true, features = ["borsh"] }
+cuprate-p2p-bucket  = { workspace = true }
 
 tower = { workspace = true, features = ["util"] }
 tokio = { workspace = true, features = ["time", "fs", "rt"]}
@@ -19,7 +20,7 @@ futures = { workspace = true, features = ["std"] }
 
 thiserror = { workspace = true }
 tracing = { workspace = true, features = ["std", "attributes"] }
-indexmap = { workspace = true, features = ["std"] }
+#indexmap = { workspace = true, features = ["std"] }
 
 rand = { workspace = true, features = ["std", "std_rng"] }
 

p2p/address-book/src/book.rs

@@ -357,6 +357,13 @@ impl<Z: BorshNetworkZone> AddressBook<Z> {
         self.connected_peers.insert(internal_peer_id, peer);
         Ok(())
     }
+    
+    fn add_white_peer(
+        &mut self,
+        white_peer: ZoneSpecificPeerListEntryBase<Z::Addr>
+    ) {
+        self.white_list.add_new_peer(white_peer);
+    }
 }
 
 impl<Z: BorshNetworkZone> Service<AddressBookRequest<Z>> for AddressBook<Z> {
@@ -401,6 +408,10 @@ impl<Z: BorshNetworkZone> Service<AddressBookRequest<Z>> for AddressBook<Z> {
                 self.handle_incoming_peer_list(peer_list);
                 Ok(AddressBookResponse::Ok)
             }
+            AddressBookRequest::AddWhitePeer(white_peer) =>{
+                self.add_white_peer(white_peer);
+                Ok(AddressBookResponse::Ok)
+            }
             AddressBookRequest::TakeRandomWhitePeer { height } => self
                 .take_random_white_peer(height)
                 .map(AddressBookResponse::Peer)

p2p/address-book/src/peer_list.rs

@@ -1,11 +1,11 @@
 use std::collections::{BTreeMap, HashMap, HashSet};
 
-use indexmap::IndexMap;
 use rand::prelude::*;
 
 use cuprate_constants::block::MAX_BLOCK_HEIGHT_USIZE;
 use cuprate_p2p_core::{services::ZoneSpecificPeerListEntryBase, NetZoneAddress, NetworkZone};
 use cuprate_pruning::PruningSeed;
+use cuprate_p2p_bucket::BucketMap;
 
 #[cfg(test)]
 pub(crate) mod tests;
@@ -16,7 +16,7 @@ pub(crate) mod tests;
 #[derive(Debug)]
 pub(crate) struct PeerList<Z: NetworkZone> {
     /// The peers with their peer data.
-    pub peers: IndexMap<Z::Addr, ZoneSpecificPeerListEntryBase<Z::Addr>>,
+    pub peers: BucketMap<ZoneSpecificPeerListEntryBase<Z::Addr>, 8>,
     /// An index of Pruning seed to address, so can quickly grab peers with the blocks
     /// we want.
     ///
@@ -33,7 +33,7 @@ pub(crate) struct PeerList<Z: NetworkZone> {
 impl<Z: NetworkZone> PeerList<Z> {
     /// Creates a new peer list.
     pub(crate) fn new(list: Vec<ZoneSpecificPeerListEntryBase<Z::Addr>>) -> Self {
-        let mut peers = IndexMap::with_capacity(list.len());
+        let mut peers = BucketMap::<_, 8>::new();
         let mut pruning_seeds = BTreeMap::new();
         let mut ban_ids = HashMap::with_capacity(list.len());
 
@@ -48,7 +48,7 @@ impl<Z: NetworkZone> PeerList<Z> {
                 .or_insert_with(Vec::new)
                 .push(peer.adr);
 
-            peers.insert(peer.adr, peer);
+            peers.push(peer);
         }
         Self {
             peers,
@@ -64,7 +64,7 @@ impl<Z: NetworkZone> PeerList<Z> {
 
     /// Adds a new peer to the peer list
     pub(crate) fn add_new_peer(&mut self, peer: ZoneSpecificPeerListEntryBase<Z::Addr>) {
-        if self.peers.insert(peer.adr, peer).is_none() {
+        if self.peers.push(peer).is_none() {
             #[expect(clippy::unwrap_or_default, reason = "It's more clear with this")]
             self.pruning_seeds
                 .entry(peer.pruning_seed)
@@ -94,7 +94,14 @@ impl<Z: NetworkZone> PeerList<Z> {
         // Take a random peer and see if it's in the list of must_keep_peers, if it is try again.
         // TODO: improve this
 
+        // Return early if no items
+        if self.peers.is_empty() {
+            return None
+        }
+        
         for _ in 0..3 {
+            
+            // In case we require a specific height
             if let Some(needed_height) = block_needed {
                 let (_, addresses_with_block) = self.pruning_seeds.iter().find(|(seed, _)| {
                     // TODO: factor in peer blockchain height?
@@ -110,17 +117,14 @@ impl<Z: NetworkZone> PeerList<Z> {
 
                 return self.remove_peer(&peer);
             } 
-            let len = self.len();
             
-            if len == 0 {
+            // Straightforward if we don't need a specific height.
-                return None;
+            let random_peer = self.peers.get_random(r);
-            }
+            if let Some(peer) = random_peer {
                 
-            let n = r.gen_range(0..len);
+                if !must_keep_peers.contains(&peer.adr) {
-
+                    return self.remove_peer(&peer.adr.clone())
-            let (&key, _) = self.peers.get_index(n).unwrap();
+                }
-            if !must_keep_peers.contains(&key) {
-                return self.remove_peer(&key);
             }
         }
 
@@ -132,26 +136,22 @@ impl<Z: NetworkZone> PeerList<Z> {
         r: &mut R,
         len: usize,
     ) -> Vec<ZoneSpecificPeerListEntryBase<Z::Addr>> {
-        let mut peers = self.peers.values().copied().choose_multiple(r, len);
+        (0..len)
-        // Order of the returned peers is not random, I am unsure of the impact of this, potentially allowing someone to make guesses about which peers
+            .filter_map(|_| self.peers.get_random(r).copied())
-        // were connected first.
+            .collect::<Vec<_>>()
-        // So to mitigate this shuffle the result.
-        peers.shuffle(r);
-        peers.drain(len.min(peers.len())..peers.len());
-        peers
     }
 
     /// Returns a mutable reference to a peer.
     pub(crate) fn get_peer_mut(
         &mut self,
-        peer: &Z::Addr,
+        peer_addr: &Z::Addr,
     ) -> Option<&mut ZoneSpecificPeerListEntryBase<Z::Addr>> {
-        self.peers.get_mut(peer)
+        self.peers.get_mut(peer_addr)
     }
 
     /// Returns true if the list contains this peer.
     pub(crate) fn contains_peer(&self, peer: &Z::Addr) -> bool {
-        self.peers.contains_key(peer)
+        self.peers.contains(peer)
     }
 
     /// Removes a peer from the pruning idx
@@ -197,7 +197,7 @@ impl<Z: NetworkZone> PeerList<Z> {
         &mut self,
         peer: &Z::Addr,
     ) -> Option<ZoneSpecificPeerListEntryBase<Z::Addr>> {
-        let peer_eb = self.peers.swap_remove(peer)?;
+        let peer_eb = self.peers.remove(peer)?;
         self.remove_peer_from_all_idxs(&peer_eb);
         Some(peer_eb)
     }
@@ -231,8 +231,8 @@ impl<Z: NetworkZone> PeerList<Z> {
             if removed_count >= target_removed {
                 break;
             }
-            if !must_keep_peers.contains(peer_adr) {
+            if !must_keep_peers.contains(&peer_adr) {
-                peers_to_remove.push(*peer_adr);
+                peers_to_remove.push(peer_adr);
                 removed_count += 1;
             }
         }

p2p/address-book/src/peer_list/tests.rs

@@ -68,7 +68,7 @@ fn peer_list_reduce_length() {
 #[test]
 fn peer_list_reduce_length_with_peers_we_need() {
     let mut peer_list = make_fake_peer_list(0, 500);
-    let must_keep_peers = peer_list.peers.keys().copied().collect::<HashSet<_>>();
+    let must_keep_peers = peer_list.peers.keys().collect::<HashSet<_>>();
 
     let target_len = 49;
 
@@ -93,7 +93,7 @@ fn peer_list_remove_specific_peer() {
         addrs.iter().for_each(|adr| assert_ne!(adr, &peer.adr));
     }
 
-    assert!(!peers.contains_key(&peer.adr));
+    assert!(!peers.contains(&peer.adr));
 }
 
 #[test]
@@ -170,7 +170,7 @@ fn peer_list_ban_peers() {
     let ban_id = peer.adr.ban_id();
 
     assert_eq!(peer_list.ban_ids.get(&ban_id), None);
-    for (addr, _) in peer_list.peers {
+    for addr in peer_list.peers.keys() {
         assert_ne!(addr.ban_id(), ban_id);
     }
 }

p2p/address-book/src/store.rs

@@ -86,11 +86,11 @@ mod tests {
         assert_eq!(gray_list.peers.len(), gray_list_2.peers.len());
 
         for addr in white_list.peers.keys() {
-            assert!(white_list_2.contains_peer(addr));
+            assert!(white_list_2.contains_peer(&addr));
         }
 
         for addr in gray_list.peers.keys() {
-            assert!(gray_list_2.contains_peer(addr));
+            assert!(gray_list_2.contains_peer(&addr));
         }
     }
 }

p2p/bucket/src/lib.rs

@@ -1,10 +1,10 @@
-//! Bucket data structure
+//! `BucketMap` data structure
 //!
 //! A collection data structure that discriminates its unique items and place them into "buckets".
 //!
 //! The item must implement the [`Bucketable`] trait that defines how to create the discriminant
-//! from the item type. The data structure will internally contain any item into "buckets" or vectors
+//! from the item type and its discriminated field. The data structure will internally contain any item into 
-//! of sized capacity `N` that regroup all the stored items with this specific discriminant.
+//! "buckets" or vectors of sized capacity `N` that regroup all the stored items with this specific discriminant.
 //!
 //! A practical example of this data structure is for storing `N` amount of IP discriminated by their subnets.
 //! You can store in each "buckets" corresponding to a `/16` subnet up to `N` IPs of that subnet.
@@ -12,11 +12,11 @@
 //! # Example
 //!
 //! ```
-//! use cuprate_p2p_bucket::Bucket;
+//! use cuprate_p2p_bucket::BucketMap;
 //! use std::net::Ipv4Addr;
 //!
 //! // Create a new bucket that can store at most 2 IPs in a particular `/16` subnet.
-//! let mut bucket = Bucket::<2,Ipv4Addr>::new();
+//! let mut bucket = BucketMap::<2,Ipv4Addr>::new();
 //!
 //! // Fulfill the `96.96.0.0/16` bucket.
 //! bucket.push("96.96.0.1".parse().unwrap());
@@ -39,28 +39,47 @@
 //! ```
 
 use arrayvec::{ArrayVec, CapacityError};
-use rand::random;
+use rand::Rng;
 
-use std::{collections::BTreeMap, net::Ipv4Addr};
+use std::{
+    collections::BTreeMap,
+    fmt::Debug, 
+    net::{IpAddr, SocketAddr}
+};
 
-/// A discriminant that can be computed from the type.
+/// This trait defines an item key and discriminant in
+/// a `BucketMap` and method to compute the discriminant
+/// from the key
 pub trait Bucketable: Sized + Eq + Clone {
-    /// The type of the discriminant being used in the Binary tree.
+    /// The type of the key being used to lookup for a specific element.
-    type Discriminant: Ord + AsRef<[u8]>;
+    /// It is also the type from which the discriminant is computed.
+    type Key: Sized + Eq + Clone;
 
-    /// Method that can compute the discriminant from the item.
+    /// The type of discriminant used to discriminate items together.
-    fn discriminant(&self) -> Self::Discriminant;
+    type Discriminant: Ord + Debug + Send;
+    
+    /// Method that give return the key from the item.
+    fn key(&self) -> Self::Key;
+    
+    /// Method for computing a discriminant from the key of the item.
+    fn compute_discriminant(key: &Self::Key) -> Self::Discriminant;
+    
+    /// Method that can compute the discriminant directly from the item.
+    fn discriminant(&self) -> Self::Discriminant {
+        Self::compute_discriminant(&self.key())
+    }
 }
 
 /// A collection data structure discriminating its unique items
 /// with a specified method. Limiting the amount of items stored
 /// with that discriminant to the const `N`.
-pub struct Bucket<const N: usize, I: Bucketable> {
+#[derive(Debug)]
+pub struct BucketMap<I: Bucketable, const N: usize> {
     /// The storage of the bucket
     storage: BTreeMap<I::Discriminant, ArrayVec<I, N>>,
 }
 
-impl<const N: usize, I: Bucketable> Bucket<N, I> {
+impl<I: Bucketable, const N: usize> BucketMap<I, N> {
     /// Create a new Bucket
     pub const fn new() -> Self {
         Self {
@@ -81,10 +100,10 @@ impl<const N: usize, I: Bucketable> Bucket<N, I> {
     /// # Example
     ///
     /// ```
-    /// use cuprate_p2p_bucket::Bucket;
+    /// use cuprate_p2p_bucket::BucketMap;
     /// use std::net::Ipv4Addr;
     ///
-    /// let mut bucket = Bucket::<8,Ipv4Addr>::new();
+    /// let mut bucket = BucketMap::<8,Ipv4Addr>::new();
     ///
     /// // Push a first IP address.
     /// bucket.push("127.0.0.1".parse().unwrap());
@@ -112,17 +131,63 @@ impl<const N: usize, I: Bucketable> Bucket<N, I> {
         None
     }
     
-    /// Will attempt to remove an item from the bucket.
+    /// Get a reference to an item stored in the `BucketMap` with 
-    pub fn remove(&mut self, item: &I) -> Option<I> {
+    /// the specified item key.
-        self.storage.get_mut(&item.discriminant()).and_then(|vec| {
+    pub fn get(&self, key: &I::Key) -> Option<&I> {
+        self.storage.get(&I::compute_discriminant(key)).and_then(|vec| {
             vec.iter()
                 .enumerate()
-                .find_map(|(i, v)| (item == v).then_some(i))
+                .find_map(|(i, v)| (*key == v.key()).then_some(i))
+                .and_then(|index| vec.get(index))
+        })
+    }
+    
+    /// Get a mutable reference to an item stored in the `BucketMap`
+    /// with the specified item key.
+    pub fn get_mut(&mut self, key: &I::Key) -> Option<&mut I> {
+        self.storage.get_mut(&I::compute_discriminant(key)).and_then(|vec| {
+            vec.iter()
+                .enumerate()
+                .find_map(|(i, v)| (*key == v.key()).then_some(i))
+                .and_then(|index| vec.get_mut(index))
+        })
+    }
+    
+    /// Return a reference to an item chosen at random.
+    ///
+    /// Repeated use of this function will provide a normal distribution of
+    /// items based on their discriminants.
+    pub fn get_random<R: Rng>(&self, r: &mut R) -> Option<&I> {
+        // Get the total amount of discriminants to explore.
+        let len = self.storage.len();
+
+        // Get a random bucket.
+        let (_, vec) = self.storage.iter().nth(r.r#gen::<usize>() / len).unwrap();
+
+        // Return a reference chose at random.
+        vec.get(r.r#gen::<usize>() / vec.len())
+    }
+    
+    /// Will attempt to remove an item from the bucket.
+    /// 
+    /// Internally use `swap_remove` which breaks the order in which
+    /// elements in a bucket were inserted.
+    pub fn remove(&mut self, key: &I::Key) -> Option<I> {
+        self.storage.get_mut(&I::compute_discriminant(key)).and_then(|vec| {
+            vec.iter()
+                .enumerate()
+                .find_map(|(i, v)| (*key == v.key()).then_some(i))
                 .map(|index| vec.swap_remove(index))
         })
     }
     
-    /// Return the number of item stored within the storage
+    /// Return `true` if the `BucketMap` contains an item
+    /// with the key `key`. Return `false` otherwise.
+    pub fn contains(&self, key: &I::Key) -> bool {
+        self.get(key).is_some()
+    }
+
+    /// Return the number of item stored within the storage.
     pub fn len(&self) -> usize {
         self.storage.values().map(ArrayVec::len).sum()
     }
@@ -135,38 +200,82 @@ impl<const N: usize, I: Bucketable> Bucket<N, I> {
         self.storage.get(discriminant).map(ArrayVec::len)
     }
 
-    /// Return `true` if the storage contains no items
+    /// Return `true` if the storage contains no items.
     pub fn is_empty(&self) -> bool {
         self.len() == 0
     }
     
-    /// Return a reference to an item chosen at random.
+    /// Return an iterator of the keys of all items contained in
-    ///
+    /// the `BucketMap`.
-    /// Repeated use of this function will provide a normal distribution of
+    pub fn keys(&self) -> impl Iterator<Item = I::Key> + '_ {
-    /// items based on their discriminants.
-    pub fn get_random(&mut self) -> Option<&I> {
-        // Get the total amount of discriminants to explore.
-        let len = self.storage.len();
         
-        // Get a random bucket.
+        self.storage
-        let (_, vec) = self.storage.iter().nth(random::<usize>() / len).unwrap();
+            .iter()
+            .flat_map(|bucket| 
+                bucket.1
+                    .iter()
+                    .map(I::key)
+            )
+    }
     
-        // Return a reference chose at random.
+    /// Return an iterator of all items contained in
-        vec.get(random::<usize>() / vec.len())
+    /// the `BucketMap`.
+    pub fn values(&self) -> impl Iterator<Item = &I> + '_ {
+        
+        self.storage
+            .iter()
+            .flat_map(|bucket| 
+                bucket.1
+                    .iter()
+            )
     }
 }
 
-impl<const N: usize, I: Bucketable> Default for Bucket<N, I> {
+impl<I: Bucketable, const N: usize> Default for BucketMap<I, N> {
     fn default() -> Self {
         Self::new()
     }
 }
 
-impl Bucketable for Ipv4Addr {
+/// A discriminant type for `SocketAddr`.
-    /// We are discriminating by `/16` subnets.
+#[derive(Debug, PartialEq, PartialOrd, Eq, Ord)]
-    type Discriminant = [u8; 2];
+pub enum IpSubnet {
+    /// The discriminant for IPv4 subnets is the 16 bytes prefix (therefore discriminating `/16` subnets)
+    V4([u8; 2]),
+    /// The equivalent of a `/16` IPv4 subnet for IPv6 is a `/32` prefix
+    V6([u8; 4])
+}
 
-    fn discriminant(&self) -> Self::Discriminant {
+impl Bucketable for SocketAddr {
-        [self.octets()[0], self.octets()[1]]
+    
+    /// The key is the IP address itself.
+    type Key = IpAddr;
+    
+    /// The discriminant is either a truncated `/16` IPv4 or a `/32` IPv6 prefix.
+    type Discriminant = IpSubnet;
+    
+    #[inline]
+    fn key(&self) -> Self::Key {
+        self.ip()
+    }
+    
+    #[inline]
+    fn compute_discriminant(key: &Self::Key) -> Self::Discriminant {
+        match key {
+            IpAddr::V4(ip_addr_v4) => {
+                let bytes = ip_addr_v4.octets();
+                IpSubnet::V4([bytes[0],bytes[1]])
+                
+            },
+            IpAddr::V6(ip_addr_v6) => {
+                let bytes = ip_addr_v6.octets();
+                IpSubnet::V6([bytes[0],bytes[1],bytes[2],bytes[3]])
+            }
+        }
     }
 }
+
+#[cfg(test)]
+mod test {
+    
+}

p2p/p2p-core/Cargo.toml

@@ -10,10 +10,11 @@ default = ["borsh"]
 borsh = ["dep:borsh", "cuprate-pruning/borsh"]
 
 [dependencies]
-cuprate-helper  = { workspace = true, features = ["asynch"], default-features = false }
+cuprate-helper      = { workspace = true, features = ["asynch"], default-features = false }
-cuprate-wire    = { workspace = true, features = ["tracing"] }
+cuprate-wire        = { workspace = true, features = ["tracing"] }
-cuprate-pruning = { workspace = true }
+cuprate-pruning     = { workspace = true }
-cuprate-types   = { workspace = true }
+cuprate-types       = { workspace = true }
+cuprate-p2p-bucket  = { workspace = true }
 
 tokio = { workspace = true, features = ["net", "sync", "macros", "time", "rt", "rt-multi-thread"]}
 tokio-util = { workspace = true, features = ["codec"] }

p2p/p2p-core/src/client/handshaker/builder/dummy.rs

@@ -102,7 +102,9 @@ impl<N: NetworkZone> Service<AddressBookRequest<N>> for DummyAddressBook {
             | AddressBookRequest::TakeRandomWhitePeer { .. } => {
                 return ready(Err("dummy address book does not hold peers".into()));
             }
-            AddressBookRequest::NewConnection { .. } | AddressBookRequest::IncomingPeerList(_) => {
+            AddressBookRequest::NewConnection { .. } 
+            | AddressBookRequest::AddWhitePeer(_)
+            | AddressBookRequest::IncomingPeerList(_) => {
                 AddressBookResponse::Ok
             }
             AddressBookRequest::GetBan(_) => AddressBookResponse::GetBan {

p2p/p2p-core/src/lib.rs

@@ -68,6 +68,7 @@ cfg_if::cfg_if! {
 
 use std::{fmt::Debug, hash::Hash};
 
+use cuprate_p2p_bucket::Bucketable;
 use futures::{Sink, Stream};
 
 use cuprate_wire::{
@@ -105,6 +106,7 @@ pub enum ConnectionDirection {
 pub trait NetZoneAddress:
     TryFrom<NetworkAddress, Error = NetworkAddressIncorrectZone>
     + Into<NetworkAddress>
+    + Bucketable
     + std::fmt::Display
     + Hash
     + Eq

p2p/p2p-core/src/services.rs

@@ -1,5 +1,6 @@
 use std::time::Instant;
 
+use cuprate_p2p_bucket::Bucketable;
 use cuprate_pruning::{PruningError, PruningSeed};
 use cuprate_wire::{CoreSyncData, PeerListEntryBase};
 
@@ -34,6 +35,23 @@ pub struct ZoneSpecificPeerListEntryBase<A: NetZoneAddress> {
     pub rpc_credits_per_hash: u32,
 }
 
+impl<A: NetZoneAddress> Bucketable for ZoneSpecificPeerListEntryBase<A> 
+{
+    type Key = A;
+    
+    type Discriminant = A::Discriminant;
+
+    #[inline]
+    fn key(&self) -> Self::Key {
+        self.adr
+    }
+    
+    #[inline]
+    fn compute_discriminant(key: &Self::Key) -> Self::Discriminant {
+        A::compute_discriminant(&key.key())
+    }
+}
+
 impl<A: NetZoneAddress> From<ZoneSpecificPeerListEntryBase<A>> for PeerListEntryBase {
     fn from(value: ZoneSpecificPeerListEntryBase<A>) -> Self {
         Self {
@@ -94,6 +112,9 @@ pub enum AddressBookRequest<Z: NetworkZone> {
     /// Tells the address book about a peer list received from a peer.
     IncomingPeerList(Vec<ZoneSpecificPeerListEntryBase<Z::Addr>>),
     
+    /// Add a new specific white peer to the peer list.
+    AddWhitePeer(ZoneSpecificPeerListEntryBase<Z::Addr>),
+
     /// Takes a random white peer from the peer list. If height is specified
     /// then the peer list should retrieve a peer that should have a full
     /// block at that height according to it's pruning seed
@@ -141,6 +162,7 @@ pub enum AddressBookResponse<Z: NetworkZone> {
     /// Response to:
     /// - [`AddressBookRequest::NewConnection`]
     /// - [`AddressBookRequest::IncomingPeerList`]
+    /// - [`AddressBookRequest::AddWhitePeer`]
     Ok,
 
     /// Response to:

p2p/p2p/src/lib.rs

@@ -2,9 +2,10 @@
 //!
 //! This crate contains a [`NetworkInterface`] which allows interacting with the Monero P2P network on
 //! a certain [`NetworkZone`]
-use std::sync::Arc;
+use std::{sync::Arc, time::Duration};
 
 use futures::FutureExt;
+use peer_pinger::PeerPinger;
 use tokio::{sync::mpsc, task::JoinSet};
 use tower::{buffer::Buffer, util::BoxCloneService, Service, ServiceExt};
 use tracing::{instrument, Instrument, Span};
@@ -23,6 +24,7 @@ pub mod config;
 pub mod connection_maintainer;
 pub mod constants;
 mod inbound_server;
+mod peer_pinger;
 
 use block_downloader::{BlockBatch, BlockDownloaderConfig, ChainSvcRequest, ChainSvcResponse};
 pub use broadcast::{BroadcastRequest, BroadcastSvc};
@@ -54,7 +56,7 @@ where
         cuprate_address_book::init_address_book(config.address_book_config.clone()).await?;
     let address_book = Buffer::new(
         address_book,
-        config.max_inbound_connections + config.outbound_connections,
+        config.max_inbound_connections + config.outbound_connections + 1,
     );
 
     // Use the default config. Changing the defaults affects tx fluff times, which could affect D++ so for now don't allow changing
@@ -96,6 +98,12 @@ where
         outbound_connector,
     );
     
+    // TODO: Export duration into P2PConfig ?
+    let peer_pinger = PeerPinger::new(
+        address_book.clone(), 
+        Duration::from_secs(30)
+    );
+
     let mut background_tasks = JoinSet::new();
 
     background_tasks.spawn(
@@ -103,6 +111,11 @@ where
             .run()
             .instrument(Span::current()),
     );
+    background_tasks.spawn(
+        peer_pinger
+            .run()
+            .instrument(Span::current()),
+    );
     background_tasks.spawn(
         inbound_server::inbound_server(
             Arc::clone(&client_pool),

p2p/p2p/src/peer_pinger.rs

@@ -0,0 +1,72 @@
+//! Peer Pinger.
+//! 
+//! This module handles the connectivity check to the peers contain in our address book.
+//! It make sure to periodically remove any peers that are unreachable.
+
+use std::{marker::PhantomData, time::Duration};
+
+use cuprate_p2p_core::{client::handshaker::ping, services::{AddressBookRequest, AddressBookResponse}, AddressBook, NetworkZone};
+use tokio::time::{sleep, timeout};
+use tracing::warn;
+
+use crate::constants::PING_REQUEST_TIMEOUT;
+
+pub(crate) struct PeerPinger<N: NetworkZone, A: AddressBook<N>> {
+    address_book_svc: A,
+    delay_btw_ping: Duration,
+    _network: PhantomData<N>
+}
+
+impl<
+    N: NetworkZone,
+    A: AddressBook<N>
+> PeerPinger<N,A> {
+    pub(crate) const fn new(address_book_svc: A, delay_btw_ping: Duration) -> Self {
+        Self {
+            address_book_svc,
+            delay_btw_ping,
+            _network: PhantomData,
+        }
+    }
+    
+    /// Will ping a random white and gray peer every `self.delay_btw_ping`.
+    /// Only replace the peer back in the list if it has been reached.
+    pub(crate) async fn run(mut self) {
+        
+        loop {
+            sleep(self.delay_btw_ping).await;
+            
+            // First ping a white peer
+            let Ok(AddressBookResponse::Peer(peer)) = self.address_book_svc.call(AddressBookRequest::TakeRandomWhitePeer { height: None }).await else {
+                warn!("AddressBook unavailable.");
+                return
+            };
+            
+            let response = timeout(PING_REQUEST_TIMEOUT, ping::<N>(peer.adr)).await;
+            if let Ok(Ok(peer_id)) = response {
+                if peer_id == peer.id {
+                    let Ok(AddressBookResponse::Ok) = self.address_book_svc.call(AddressBookRequest::AddWhitePeer(peer)).await else {
+                        warn!("AddressBook unavailable.");
+                        return
+                    };
+                }
+            }
+            
+            // Then ping a gray peer
+            let Ok(AddressBookResponse::Peer(peer)) = self.address_book_svc.call(AddressBookRequest::TakeRandomGrayPeer { height: None }).await else {
+                warn!("AddressBook unavailable.");
+                return
+            };
+            
+            let response = timeout(PING_REQUEST_TIMEOUT, ping::<N>(peer.adr)).await;
+            if let Ok(Ok(peer_id)) = response {
+                if peer_id == peer.id {
+                    let Ok(AddressBookResponse::Ok) = self.address_book_svc.call(AddressBookRequest::IncomingPeerList(vec![peer])).await else {
+                        warn!("AddressBook unavailable.");
+                        return
+                    };
+                }
+            }
+        }
+    }
+}

test-utils/Cargo.toml

@@ -10,6 +10,7 @@ cuprate-types             = { workspace = true }
 cuprate-helper            = { workspace = true, features = ["map", "tx"] }
 cuprate-wire              = { workspace = true }
 cuprate-p2p-core          = { workspace = true, features = ["borsh"] }
+cuprate-p2p-bucket        = { workspace = true }
 
 hex                       = { workspace = true }
 hex-literal               = { workspace = true }

test-utils/src/test_netzone.rs

@@ -11,6 +11,7 @@ use std::{
 };
 
 use borsh::{BorshDeserialize, BorshSerialize};
+use cuprate_p2p_bucket::Bucketable;
 use futures::Stream;
 use tokio::io::{DuplexStream, ReadHalf, WriteHalf};
 use tokio_util::codec::{FramedRead, FramedWrite};
@@ -23,9 +24,23 @@ use cuprate_wire::{
 use cuprate_p2p_core::{NetZoneAddress, NetworkZone};
 
 /// An address on the test network
-#[derive(Debug, Clone, Copy, Eq, Hash, PartialEq, BorshSerialize, BorshDeserialize)]
+#[derive(Debug, Clone, Copy, Eq, Hash, PartialEq, PartialOrd, Ord, BorshSerialize, BorshDeserialize)]
 pub struct TestNetZoneAddr(pub u32);
 
+impl Bucketable for TestNetZoneAddr {
+    type Key = Self;
+
+    type Discriminant = Self;
+
+    fn key(&self) -> Self::Key {
+        *self
+    }
+
+    fn compute_discriminant(key: &Self::Key) -> Self::Discriminant {
+        *key
+    }
+}
+
 impl NetZoneAddress for TestNetZoneAddr {
     type BanID = Self;