tiborkiss commented on issue #3044: URL: https://github.com/apache/iceberg/issues/3044#issuecomment-907758924
Last one-two days I wrote a streaming job and packaged as a fat jar. I had to add the minimal hadoop dependency and in the code I had to initialize the hadoop conf with the constructor which has a boolean false, telling that to not load any hadoop config file (since I don't have in Flink). Finally, I managed to work with the streaming job, which is running in Flink 1.13.1 (behind Ververica CE). Definitively, I see a good reason to remove the hadoop conf dependency from GlueCatalog, which uses the custom catalog from iceberg code. Currently there are three catalog implementation classes and somehow GlueCatalog falls under the custom catalog implementation. From design aspects, the constellation of these three type of catalogs and how another specialized like GlueCatalog is used, may cause some confusion at first time when you try to use. Some refactoring, such as removal of hadoop conf or more, it would be nice to see. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
