Log stuck lookups (#5778)

* Log stuck lookups every interval

* Implement debug manually

* Add comment

* Do not print peers twice

* Add SYNC_LOOKUPS_STUCK metric

* Skip logging request root

* use derivative

* Merge branch 'unstable' of https://github.com/sigp/lighthouse into log-stuck-lookups

* add req id to debug

* Merge remote-tracking branch 'sigp/unstable' into log-stuck-lookups

* Fix conflict with unstable

beacon_node/network/src/sync/manager.rs

@@ -547,6 +547,10 @@ impl<T: BeaconChainTypes> SyncManager<T> {
             futures::stream::iter(ee_responsiveness_watch.await).flatten()
         };
 
+        // LOOKUP_MAX_DURATION_SECS is 60 seconds. Logging every 30 seconds allows enough timely
+        // visbility while being sparse and not increasing the debug log volume in a noticeable way
+        let mut interval = tokio::time::interval(Duration::from_secs(30));
+
         // process any inbound messages
         loop {
             tokio::select! {
@@ -556,6 +560,9 @@ impl<T: BeaconChainTypes> SyncManager<T> {
                 Some(engine_state) = check_ee_stream.next(), if check_ee => {
                     self.handle_new_execution_engine_state(engine_state);
                 }
+                _ = interval.tick() => {
+                    self.block_lookups.log_stuck_lookups();
+                }
             }
         }
     }