Related to the post: Migration from DynamoDB to Spanner/BigTable
I have another challenge here. How to create the schema for all 'N' tables in Spanner? Is there any utility available for this? There are TBs worth of data in DynamoDB. Not sure how to analyse that for schema info.
Unfortunately, there is no official migration tool. Translating NoSQL to relational schema without "understanding" the data is a best guess scenario at best.
According to documentation (aws-dynamodb) you can use describe table to get the current schema for each dynamo db table and how to map it to Cloud Spanner.
Use your understanding of the data to map interleaved tables, choose optimal primary keys and secondary indexes, this would give you better performance than naive 1-1 migration.
You can then follow the guide listed here
Use the following for schema design on Cloud Spanner: doc
I hope this helps.
