FrankChen021 commented on PR #20183: URL: https://github.com/apache/druid/pull/20183#issuecomment-6365071107
> At a high level the approach looks OK, although I wonder why you used `/druid/v2/` for fanout rather than the existing APIs? I What do you mean by existing APIs here? do you mean the API for the server properties like `/status/properties` ? The reason is that I would like the internal data fetching from different servers to be unified like how we execute sub queries on different nodes. By this way, in future, there's any new table(like the sys.stack_trace table) added, we don't add new APIs. Currently SCAN query is fan out from broker to different nodes, and I was planning to push down aggregation queries in a later stage. So the /druid/v2 is the best place to serve the need. However, for the native engine, I will not do the aggregation query push down, because I found that the the dart engine already does this, so the long term evolution here is to use the dart engine to fan out sub queries. -- 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]
