codope commented on PR #5449: URL: https://github.com/apache/hudi/pull/5449#issuecomment-1113976714
> @codope I read through the blog. Thanks for the great writeup. I have a couple of thoughts/questions from an external perspective. > > * Should we name this blog to explicitly mention as Asynchronous metadata indexing using Hudi. This will avoid confusion with the current indexing concept here for users - https://hudi.apache.org/docs/indexing ? > * Also I see that we have a category of `Services` in our Docs. There is a page for clustering, compaction, etc in there. Since this is one of the table services, should we also add a page on metadata indexing under `Services`. It can have a brief description and point to this blog for detailed content. > * One last thing is, seeing the description, I also wanted to understand what happens if we disable `hoodie.metadata.index.async` . What is the behavior then ? Does it run inline ? What are the implications. It would be good to see those explained as well. > > Let me know your thoughts. Good points @bhasudha I have moved the setup and usage related sections to a separate page under `Services` tab, check out #5476 I need to do some more work on this blog. I will rename, add more design elements and diagram. If async indexing is disabled with any of the index types enabled then those metadata partitions will be created/updated synchronously. Added this line in caveat section in #5476 -- 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]
