Add PeerDAS metrics to track subnets without peers (#6928)

Currently we track a key metric `PEERS_PER_COLUMN_SUBNET` in a getter `good_peers_on_sampling_subnets`. Another PR https://github.com/sigp/lighthouse/pull/6922 deletes that function, so we have to move the metric anyway. This PR moves that metric computation to the metrics spawned task which is refreshed every 5 seconds.

I also added a few more useful metrics. The total set and intended usage is:

- `sync_peers_per_column_subnet`: Track health of overall subnets in your node
- `sync_peers_per_custody_column_subnet`: Track health of the subnets your node needs. We should track this metric closely in our dashboards with a heatmap and bar plot
- ~~`sync_column_subnets_with_zero_peers`: Is equivalent to the Grafana query `count(sync_peers_per_column_subnet == 0) by (instance)`. We may prefer to skip it, but I believe it's the most important metric as if `sync_column_subnets_with_zero_peers > 0` your node stalls.~~
- ~~`sync_custody_column_subnets_with_zero_peers`: `count(sync_peers_per_custody_column_subnet == 0) by (instance)`~~

beacon_node/lighthouse_network/src/peer_manager/peerdb/peer_info.rs

@@ -234,6 +234,11 @@ impl<E: EthSpec> PeerInfo<E> {
         self.custody_subnets.contains(subnet)
     }
 
+    /// Returns an iterator on this peer's custody subnets
+    pub fn custody_subnets_iter(&self) -> impl Iterator<Item = &DataColumnSubnetId> {
+        self.custody_subnets.iter()
+    }
+
     /// Returns true if the peer is connected to a long-lived subnet.
     pub fn has_long_lived_subnet(&self) -> bool {
         // Check the meta_data

beacon_node/network/src/metrics.rs

@@ -14,6 +14,7 @@ pub use metrics::*;
 use std::sync::{Arc, LazyLock};
 use strum::AsRefStr;
 use strum::IntoEnumIterator;
+use types::DataColumnSubnetId;
 use types::EthSpec;
 
 pub const SUCCESS: &str = "SUCCESS";
@@ -374,11 +375,18 @@ pub static PEERS_PER_SYNC_TYPE: LazyLock<Result<IntGaugeVec>> = LazyLock::new(||
 });
 pub static PEERS_PER_COLUMN_SUBNET: LazyLock<Result<IntGaugeVec>> = LazyLock::new(|| {
     try_create_int_gauge_vec(
-        "peers_per_column_subnet",
+        "sync_peers_per_column_subnet",
         "Number of connected peers per column subnet",
         &["subnet_id"],
     )
 });
+pub static PEERS_PER_CUSTODY_COLUMN_SUBNET: LazyLock<Result<IntGaugeVec>> = LazyLock::new(|| {
+    try_create_int_gauge_vec(
+        "sync_peers_per_custody_column_subnet",
+        "Number of connected peers per custody column subnet",
+        &["subnet_id"],
+    )
+});
 pub static SYNCING_CHAINS_COUNT: LazyLock<Result<IntGaugeVec>> = LazyLock::new(|| {
     try_create_int_gauge_vec(
         "sync_range_chains",
@@ -746,16 +754,42 @@ pub fn update_sync_metrics<E: EthSpec>(network_globals: &Arc<NetworkGlobals<E>>)
 
     // count per sync status, the number of connected peers
     let mut peers_per_sync_type = FnvHashMap::default();
-    for sync_type in network_globals
+    let mut peers_per_column_subnet = FnvHashMap::default();
-        .peers
+
-        .read()
+    for (_, info) in network_globals.peers.read().connected_peers() {
-        .connected_peers()
+        let sync_type = info.sync_status().as_str();
-        .map(|(_peer_id, info)| info.sync_status().as_str())
-    {
         *peers_per_sync_type.entry(sync_type).or_default() += 1;
+
+        for subnet in info.custody_subnets_iter() {
+            *peers_per_column_subnet.entry(*subnet).or_default() += 1;
+        }
     }
 
     for (sync_type, peer_count) in peers_per_sync_type {
         set_gauge_entry(&PEERS_PER_SYNC_TYPE, &[sync_type], peer_count);
     }
+
+    let all_column_subnets =
+        (0..network_globals.spec.data_column_sidecar_subnet_count).map(DataColumnSubnetId::new);
+    let custody_column_subnets = network_globals.sampling_subnets.iter();
+
+    // Iterate all subnet values to set to zero the empty entries in peers_per_column_subnet
+    for subnet in all_column_subnets {
+        set_gauge_entry(
+            &PEERS_PER_COLUMN_SUBNET,
+            &[&format!("{subnet}")],
+            peers_per_column_subnet.get(&subnet).copied().unwrap_or(0),
+        );
+    }
+
+    // Registering this metric is a duplicate for supernodes but helpful for fullnodes. This way
+    // operators can monitor the health of only the subnets of their interest without complex
+    // Grafana queries.
+    for subnet in custody_column_subnets {
+        set_gauge_entry(
+            &PEERS_PER_CUSTODY_COLUMN_SUBNET,
+            &[&format!("{subnet}")],
+            peers_per_column_subnet.get(subnet).copied().unwrap_or(0),
+        );
+    }
 }

beacon_node/network/src/sync/range_sync/chain.rs

@@ -1,7 +1,6 @@
 use super::batch::{BatchInfo, BatchProcessingResult, BatchState};
 use super::RangeSyncType;
 use crate::metrics;
-use crate::metrics::PEERS_PER_COLUMN_SUBNET;
 use crate::network_beacon_processor::ChainSegmentProcessId;
 use crate::sync::network_context::RangeRequestId;
 use crate::sync::{network_context::SyncNetworkContext, BatchOperationOutcome, BatchProcessResult};
@@ -10,7 +9,6 @@ use beacon_chain::BeaconChainTypes;
 use fnv::FnvHashMap;
 use lighthouse_network::service::api_types::Id;
 use lighthouse_network::{PeerAction, PeerId};
-use metrics::set_int_gauge;
 use rand::seq::SliceRandom;
 use rand::Rng;
 use slog::{crit, debug, o, warn};
@@ -1106,11 +1104,6 @@ impl<T: BeaconChainTypes> SyncingChain<T> {
                         .good_custody_subnet_peer(*subnet_id)
                         .count();
 
-                    set_int_gauge(
-                        &PEERS_PER_COLUMN_SUBNET,
-                        &[&subnet_id.to_string()],
-                        peer_count as i64,
-                    );
                     peer_count > 0
                 });
             peers_on_all_custody_subnets