diff --git a/crates/federate/src/lib.rs b/crates/federate/src/lib.rs
index 0c88c3a6e..c72e51015 100644
--- a/crates/federate/src/lib.rs
+++ b/crates/federate/src/lib.rs
@@ -158,6 +158,7 @@ async fn print_stats(pool: &mut DbPool<'_>, stats: &HashMap<String, FederationQu
   );
   // todo: less noisy output (only output failing instances and summary for successful)
   // todo: more stats (act/sec, avg http req duration)
+  let mut ok_count = 0;
   for stat in stats.values() {
     let behind = last_id - stat.last_successful_id;
     if stat.fail_count > 0 {
@@ -169,7 +170,12 @@ async fn print_stats(pool: &mut DbPool<'_>, stats: &HashMap<String, FederationQu
         retry_sleep_duration(stat.fail_count)
       );
     } else {
-      tracing::info!("{}: Ok. {} behind", stat.domain, behind);
+      if behind > 0 {
+        tracing::info!("{}: Ok. {} behind", stat.domain, behind);
+      } else {
+        ok_count += 1;
+      }
     }
   }
+  tracing::info!("{ok_count} others up to date");
 }
diff --git a/crates/federate/src/worker.rs b/crates/federate/src/worker.rs
index 5c9910e14..240b37d65 100644
--- a/crates/federate/src/worker.rs
+++ b/crates/federate/src/worker.rs
@@ -107,14 +107,19 @@ pub async fn instance_worker(
         let mut req = sign_raw(&task, &data, REQWEST_TIMEOUT).await?;
         tracing::info!("sending out {}", task);
         while let Err(e) = send_raw(&task, &data, req).await {
-          tracing::info!("{task} failed: {e}");
           state.fail_count += 1;
           state.last_retry = Utc::now();
+          let retry_delay: Duration = retry_sleep_duration(state.fail_count);
+          tracing::info!(
+            "{}: retrying {id} attempt {} with delay {retry_delay:.2?}. ({e})",
+            instance.domain,
+            state.fail_count
+          );
           stats_sender.send(state.clone())?;
           FederationQueueState::upsert(&mut pool, &state).await?;
           req = sign_raw(&task, &data, REQWEST_TIMEOUT).await?; // resign request
           tokio::select! {
-            () = sleep(retry_sleep_duration(state.fail_count)) => {},
+            () = sleep(retry_delay) => {},
             () = stop.cancelled() => {
               // save state to db and exit
               break 'batch;