The MySQL server is running with the --read-only

[Ojas Desai]

Setup:
3 node MySQL 5.7.21 Innodb cluster community Edition, single primary.
MySQL Router: 2.1.4

We have been facing scenarios where in application pools 20 connections onto the cluster.
No when the primary switched over to a different node (due to network fluctuations) and the old primary is secondary now.

Now the connections which were established on the primary earlier do not move to the new primary or get killed/terminated. The application throws the below error as the node is a secondary.

Caused by: java.sql.BatchUpdateException: The MySQL server is running with the
--read-only option so it cannot execute this statement

Is this something MySQLRouter/Innodb cluster needs to handle ?
or any configuration to be take care off or any feature ?