nexus: allow instance updates to complete when one switch is down

nexus/src/app/instance_network.rs

@@ -652,14 +652,31 @@ pub(crate) async fn instance_delete_dpd_config(
         instance_id,
     )
     .await;
-    notify_dendrite_nat_state(
+
+    // Just like in `instance_ensure_dpd_config`, we should not bail out if
+    // there is an error while notifying dendrite. If there is an error
+    // communicating with one dendrite instance but the other is operational and
+    // we bail here, it will prevent Nexus from tearing down the VMM's network
+    // config, leaving the instance stuck and preventing it from being restarted
+    // or deleted. Dendrite should still catch back up via a RPW if we fail to
+    // notify it here.
+    if let Err(error) = notify_dendrite_nat_state(
         datastore,
         log,
         resolver,
         opctx_alloc,
         Some(instance_id),
     )
     .await
+    {
+        warn!(
+            log,
+            "error encountered while notifying dendrite of deleted instance";
+            "error" => InlineErrorChain::new(&error),
+        );
+    }
+
+    Ok(())
 }
 
 async fn instance_send_attached_subnets_to_dendrite(