cloud-fan commented on issue #27722: [SPARK-30969][CORE] Remove resource coordination support from Standalone URL: https://github.com/apache/spark/pull/27722#issuecomment-592071235 I'm not very familiar with the whole story, but the recent discussion of backward compatibility makes me worry about this one. If we release a very complicated API for 1% of the user cases, it may not worth the cost to maintain that API. Already implemented/tested doesn't count the future maintenance cost. Recently I'm suffered from keeping the support of some legacy behavior/API and I hope we can avoid such problems at the beginning. Are there any valid use cases for multi-worker per host? If not or only a little, I'd suggest we at least don't support it in the new GPU scheduling.
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
