The warnings in the tracelog related to memory limits are unrelated - it can have an effect on replication performance, but it doesn’t affect correctness.
From looking at your original tracelogs, it seems like the DR replicas are trying to connect to ‘127.0.0.1’ or ‘localhost’. How did you setup your primary cluster? What’s the output of SHOW LEAVES and SHOW AGGREGATORS?
From the tracelogs it seems like you setup your primary cluster with non-public IPs (bootstrap aggregators '127.0.0.1' and/or add leaf root:PW@'127.0.0.1':3307). That does not work with DR - the address you provide in the REPLICATE command is only used for initial setup, and then the addresses used when setting up the primary cluster are used to actually replicate the data (this is because the leaves in the secondary cluster connect directly to the primary cluster’s leaves). The addresses don’t need to be public, but need to be correctly routable to from the secondary cluster.