Which database system?: target is Snowflake on AWS

Additional details:

Can you make resilient work on TimeKey channels (in Append Mode) to detect duplicates? Documentation says "A row's existence is determined using the replication key known to HVR". What we've already discussed in support cases is the fact that our source system hardly provides any primary key. However, isn't the hvr_tx_seq - based on source rrn per journal receiver - a possibility to detect duplications?