network:bridge: fix peer_count metric (paritytech#3711)

The metric records the current protocol_version of the validator that
just connected with the peer_map.len(), which contains all peers that
connected, that has the effect the metric will be wrong since it won't
tell us how many peers we have connected per version because it will
always record the total number of peers

Fix this by counting by version inside peer_map, additionally because
that might be a bit heavier than len(), publish it only on-active
leaves.

---------

Signed-off-by: Alexandru Gheorghe <alexandru.gheorghe@parity.io>

polkadot/node/network/bridge/src/lib.rs

@@ -102,6 +102,30 @@ struct SharedInner {
  collation_peers: HashMap<PeerId, PeerData>,
 }
 
+// Counts the number of peers that are connectioned using `version`
+fn count_peers_by_version(peers: &HashMap<PeerId, PeerData>) -> HashMap<ProtocolVersion, usize> {
+ let mut by_version_count = HashMap::new();
+ for peer in peers.values() {
+ *(by_version_count.entry(peer.version).or_default()) += 1;
+ }
+ by_version_count
+}
+
+// Notes the peer count
+fn note_peers_count(metrics: &Metrics, shared: &Shared) {
+ let guard = shared.0.lock();
+ let validation_stats = count_peers_by_version(&guard.validation_peers);
+ let collation_stats = count_peers_by_version(&guard.collation_peers);
+
+ for (version, count) in validation_stats {
+ metrics.note_peer_count(PeerSet::Validation, version, count)
+ }
+
+ for (version, count) in collation_stats {
+ metrics.note_peer_count(PeerSet::Collation, version, count)
+ }
+}
+
 pub(crate) enum Mode {
  Syncing(Box<dyn SyncOracle + Send>),
  Active,

polkadot/node/network/bridge/src/rx/mod.rs

@@ -262,7 +262,6 @@ async fn handle_validation_message<AD>(
  }
 
  metrics.on_peer_connected(peer_set, version);
- metrics.note_peer_count(peer_set, version, peer_map.len());
 
  shared.local_view.clone().unwrap_or(View::default())
  };
@@ -320,8 +319,6 @@ async fn handle_validation_message<AD>(
  let w = peer_map.remove(&peer).is_some();
 
  metrics.on_peer_disconnected(peer_set, version);
- metrics.note_peer_count(peer_set, version, peer_map.len());
-
  w
  };
 
@@ -524,7 +521,6 @@ async fn handle_collation_message<AD>(
  }
 
  metrics.on_peer_connected(peer_set, version);
- metrics.note_peer_count(peer_set, version, peer_map.len());
 
  shared.local_view.clone().unwrap_or(View::default())
  };
@@ -575,7 +571,6 @@ async fn handle_collation_message<AD>(
  let w = peer_map.remove(&peer).is_some();
 
  metrics.on_peer_disconnected(peer_set, version);
- metrics.note_peer_count(peer_set, version, peer_map.len());
 
  w
  };
@@ -832,6 +827,7 @@ where
  &metrics,
  &notification_sinks,
  );
+ note_peers_count(&metrics, &shared);
  }
  }
  },