Critical read error with sync forced mariadb cluster

Hello,

I'm using 10.5.11-MariaDB on Debian 10.10 with a read-write split router configuration on maxscale.

The mariadb galera cluster nodes have all syncing forced enabled with this parameter:
wsrep_sync_wait=7

That means consistency is enabled on the cluster.

Nevertheless, a critical read error still happens every day in the process of inserting an item and then reading it.

The read fails as if the nodes of the cluster were not synced with the last write.

MariaDB is not finding the just inserted row despite wsrep_sync_wait=7.