You must create a customer managed KMS key in the destination Region and then create a snapshot copy grant against that key so Redshift can re- encrypt the incoming snapshot with the destination CMK.
In the source Region you then enable cross-Region snapshot copying and reference the grant name you created in the destination Region. This tells Redshift which CMK (via the grant) to use when copying the encrypted snapshot into the DR Region.

Amazon DynamoDB offers two streaming models for change data capture,
Kinesis Data Streams for DynamoDB: Data retention with this model is 1 year. It can process stream records using AWS Lambda, Amazon Managed Service for Apache Flink, Kinesis data firehose, or AWS Glue streaming ETL. AWS Glue Streaming ETL can be used to cleanse, and transform streaming data and store it in either Amazon S3 data lakes or JDBC data stores.
DynamoDB Streams: This model retains data for 24 hours. It can process stream records using AWS Lambda and the DynamoDB Streams Kinesis adapter.
Reference:
https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/streamsmain.html

An Amazon Redshift stored procedure can perform the delete-and-insert logic directly inside the warehouse in a single controlled operation. This is the most appropriate way to remove stale rows from the target table and load the latest records from the staging table while keeping the orders table current for downstream applications, without introducing unnecessary data movement or external processing.

The "git fetch" command will pull information about all the branches in the CodeCommit repository to the local repo.
Reference:
https://docs.aws.amazon.com/codecommit/latest/userguide/how-to-basic-git.html