Prevent publisher jobs from erroring if the connection pool is full

A full pool is a soft-error. Snooze the job for 30 seconds and try again.

changelog.d/gun_pool2.fix

@@ -0,0 +1 @@
+Connection pool errors when publishing an activity is a soft-error that will be retried shortly.

lib/pleroma/web/activity_pub/publisher.ex

@@ -129,6 +129,10 @@ def publish_one(%{inbox: inbox, json: json, actor: %User{} = actor, id: id} = pa
           _ -> {:error, e}
         end
 
+      {:error, :pool_full} ->
+        Logger.debug("Publisher snoozing worker job due to full connection pool")
+        {:snooze, 30}
+
       e ->
         unless params[:unreachable_since], do: Instances.set_unreachable(inbox)
         Logger.metadata(activity: id, inbox: inbox)