Thanks, Jack and Ryan! I think it is necessary to extend metastore and provide REST API service, so that we can not rely on metastore such as hive metastore in production.
big +1 Forward Ryan Blue <b...@apache.org> 于2021年6月10日周四 上午7:52写道： > Thanks, Jack! I agree that it would be great to get an idea of what > catalogs are out there so we can add them to Iceberg. Always good to share > those implementations. And not just for DynamoDB, too, if there are any > others out there that aren't too specific to a company's internal metastore. > > One thing I've thought about recently is defining a REST API for catalogs > so that it's easier to build a service and not just directly on top of a > database. That way you'd be able to scale out the service and not overwhelm > your data store with connections. I know that's a concern for people > running Hive metastores. If there's any interest in this, let's make some > time to think about what we might want to include there. > > Ryan > > On Tue, Jun 8, 2021 at 11:25 PM Ye, Jack <yzhao...@amazon.com.invalid> > wrote: > >> Hi everyone, >> >> >> >> Because of the many asks we have received for supporting AWS DynamoDB >> Catalog implementation, I have put up PR >> https://github.com/apache/iceberg/pull/2688 showing the DynamoDbCatalog >> that we designed and implemented. >> >> >> >> I have heard a few internal implementations based on DynamoDB in >> different companies and organizations so far. It would be great if we can >> share our designs to compare the pros and cons of these implementations and >> arrive at an optimal solution. >> >> >> >> Any suggestion or comment would be greatly appreciated! >> >> >> >> Thanks, >> >> Jack Ye >> > > > -- > Ryan Blue >
