about redshift HA, what happens when the header node fails.

Amazon Redshift will automatically detect and replace a failed node in your data warehouse cluster. As to RTO, it depends on many factors and can span anywhere between minutes to an hour. AWS does not promise an SLA.

As to RPO, it depends on the size of your Redshift cluster. If this is a single-node cluster (which is NOT recommended for customer production use), there is only one copy of the data in the cluster. When it is down, AWS needs to restore the cluster from the most recent snapshot on S3 and that becomes your RPO. For a multi-node Redshift cluster, when a lead node is down, the cluster will be down. But once the failed lead node recovers, the cluster comes back up at exactly the same point as when it crashed (including all successfully committed transactions up to crash). So there will be no data loss. For those "in flight" transactions that were interrupted by the crash, you need to re-run them.