As mentioned in #164, savepoints are not currently supported for Alternator/DynamoDB. This means that if something goes wrong during a migration (e.g. an outage on the source database, Spark cluster, or target database), the migration will have to be restarted from zero, which can take a long time.
To implement it we would have to use an equivalent of CQL’s token ranges in DynamoDB.
As mentioned in #164, savepoints are not currently supported for Alternator/DynamoDB. This means that if something goes wrong during a migration (e.g. an outage on the source database, Spark cluster, or target database), the migration will have to be restarted from zero, which can take a long time.
To implement it we would have to use an equivalent of CQL’s token ranges in DynamoDB.