Replicating to spider is fragile without retries

I encountered a failure in the replication to a spider table that is not mentioned in the recommendations at https://mariadb.com/kb/en/library/spider-storage-engine-overview/#some-server-variables-to-set-when-using-spider.

From 'show slave status\G':

                Last_SQL_Errno: 1429

                Last_SQL_Error: Could not execute Update_rows_v1 event on table db.t3; Unable to connect to foreign data source: spider-datanode3, Error_code: 1429; handler error No Error!; the event's master log binlog.000164, end_log_pos 52742152

It did work again after:

start slave sql_thread;

Can these issues be retried within the spider engine instead of failing in the engine and retried in the replication? (or are they failing after multiple attempts in the engine and the replication just continues to retry until it succeeds?)

Also I would expect the default settings to handle this, rather than having to add configuration mentioned in the link above.