checkpointing for streams
Opened this issue · 2 comments
- Common approach to handle checkpointing while using GO sdks?
- How to handle missing/duplicated data?
What the hell is checkpointing? I didn't use DynamoDB for some years, is is new term there?
How to handle missing/duplicated data?
However you like, this library only wraps complexity of iterating over overengineered dynamo streams
By checkpoint, I mean tracking the status of shards/data processed.
Refer: https://docs.aws.amazon.com/streams/latest/dev/kinesis-record-processor-ddb.html
https://www.infoq.com/articles/demystifying-dynamodb-streams
KCL adapter for dynamoDB streams has checkpoint feature (only available in Java): https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/Streams.KCLAdapter.html
The library is really well written. I had few doubts regarding how data accuracy can be maintained using go SDK.