Sync peer attribution (#7733)

Which issue # does this PR address?

Closes #7604


  Improvements to range sync including:

1. Contain column requests only to peers that are part of the SyncingChain
2. Attribute the fault to the correct peer and downscore them if they don't return the data columns for the request
3. Improve sync performance by retrying only the failed columns from other peers instead of failing the entire batch
4. Uses the earliest_available_slot to make requests to peers that claim to have the epoch. Note: if no earliest_available_slot info is available, fallback to using previous logic i.e. assume peer has everything backfilled upto WS checkpoint/da boundary

Tested this on fusaka-devnet-2 with a full node and supernode and the recovering logic seems to works well.
Also tested this a little on mainnet.

Need to do more testing and possibly add some unit tests.

beacon_node/network/src/sync/tests/range.rs

@@ -77,7 +77,7 @@ impl TestRig {
     /// Produce a head peer with an advanced head
     fn add_head_peer_with_root(&mut self, head_root: Hash256) -> PeerId {
         let local_info = self.local_info();
-        self.add_random_peer(SyncInfo {
+        self.add_supernode_peer(SyncInfo {
             head_root,
             head_slot: local_info.head_slot + 1 + Slot::new(SLOT_IMPORT_TOLERANCE as u64),
             ..local_info
@@ -93,7 +93,7 @@ impl TestRig {
     fn add_finalized_peer_with_root(&mut self, finalized_root: Hash256) -> PeerId {
         let local_info = self.local_info();
         let finalized_epoch = local_info.finalized_epoch + 2;
-        self.add_random_peer(SyncInfo {
+        self.add_supernode_peer(SyncInfo {
             finalized_epoch,
             finalized_root,
             head_slot: finalized_epoch.start_slot(E::slots_per_epoch()),
@@ -132,13 +132,13 @@ impl TestRig {
         }
     }
 
-    fn add_random_peer_not_supernode(&mut self, remote_info: SyncInfo) -> PeerId {
+    fn add_fullnode_peer(&mut self, remote_info: SyncInfo) -> PeerId {
         let peer_id = self.new_connected_peer();
         self.send_sync_message(SyncMessage::AddPeer(peer_id, remote_info));
         peer_id
     }
 
-    fn add_random_peer(&mut self, remote_info: SyncInfo) -> PeerId {
+    fn add_supernode_peer(&mut self, remote_info: SyncInfo) -> PeerId {
         // Create valid peer known to network globals
         // TODO(fulu): Using supernode peers to ensure we have peer across all column
         // subnets for syncing. Should add tests connecting to full node peers.
@@ -148,17 +148,13 @@ impl TestRig {
         peer_id
     }
 
-    fn add_random_peers(&mut self, remote_info: SyncInfo, count: usize) {
-        for _ in 0..count {
+    fn add_fullnode_peers(&mut self, remote_info: SyncInfo, peer_count: usize) {
+        for _ in 0..peer_count {
             let peer = self.new_connected_peer();
-            self.add_peer(peer, remote_info.clone());
+            self.send_sync_message(SyncMessage::AddPeer(peer, remote_info.clone()));
         }
     }
 
-    fn add_peer(&mut self, peer: PeerId, remote_info: SyncInfo) {
-        self.send_sync_message(SyncMessage::AddPeer(peer, remote_info));
-    }
-
     fn assert_state(&self, state: RangeSyncType) {
         assert_eq!(
             self.sync_manager
@@ -562,19 +558,14 @@ const EXTRA_SYNCED_EPOCHS: u64 = 2 + 1;
 fn finalized_sync_enough_global_custody_peers_few_chain_peers() {
     // Run for all forks
     let mut r = TestRig::test_setup();
-    // This test creates enough global custody peers to satisfy column queries but only adds few
-    // peers to the chain
-    r.new_connected_peers_for_peerdas();
 
     let advanced_epochs: u64 = 2;
     let remote_info = r.finalized_remote_info_advanced_by(advanced_epochs.into());
 
-    // Current priorization only sends batches to idle peers, so we need enough peers for each batch
-    // TODO: Test this with a single peer in the chain, it should still work
-    r.add_random_peers(
-        remote_info,
-        (advanced_epochs + EXTRA_SYNCED_EPOCHS) as usize,
-    );
+    // Generate enough peers and supernodes to cover all custody columns
+    let peer_count = 100;
+    r.add_fullnode_peers(remote_info.clone(), peer_count);
+    r.add_supernode_peer(remote_info);
     r.assert_state(RangeSyncType::Finalized);
 
     let last_epoch = advanced_epochs + EXTRA_SYNCED_EPOCHS;
@@ -592,9 +583,9 @@ fn finalized_sync_not_enough_custody_peers_on_start() {
     let advanced_epochs: u64 = 2;
     let remote_info = r.finalized_remote_info_advanced_by(advanced_epochs.into());
 
-    // Unikely that the single peer we added has enough columns for us. Tests are determinstic and
+    // Unikely that the single peer we added has enough columns for us. Tests are deterministic and
     // this error should never be hit
-    r.add_random_peer_not_supernode(remote_info.clone());
+    r.add_fullnode_peer(remote_info.clone());
     r.assert_state(RangeSyncType::Finalized);
 
     // Because we don't have enough peers on all columns we haven't sent any request.
@@ -603,14 +594,9 @@ fn finalized_sync_not_enough_custody_peers_on_start() {
     r.expect_empty_network();
 
     // Generate enough peers and supernodes to cover all custody columns
-    r.new_connected_peers_for_peerdas();
-    // Note: not necessary to add this peers to the chain, as we draw from the global pool
-    // We still need to add enough peers to trigger batch downloads with idle peers. Same issue as
-    // the test above.
-    r.add_random_peers(
-        remote_info,
-        (advanced_epochs + EXTRA_SYNCED_EPOCHS - 1) as usize,
-    );
+    let peer_count = 100;
+    r.add_fullnode_peers(remote_info.clone(), peer_count);
+    r.add_supernode_peer(remote_info);
 
     let last_epoch = advanced_epochs + EXTRA_SYNCED_EPOCHS;
     r.complete_and_process_range_sync_until(last_epoch, filter());