You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2020/12/08 16:32:19 UTC

[GitHub] [flink] azagrebin commented on pull request #13879: [FLINK-19832][coordination] Do not schedule shared slot bulk if some slots have failed immediately

azagrebin commented on pull request #13879:
URL: https://github.com/apache/flink/pull/13879#issuecomment-740744366


   Thanks for the review @tillrohrmann.
   I agree we do not need this `Allocated` state check in `SharedSlot.allocateLogicalSlot`.
   I decided to try removing it and splitting the shared and logical slot allocation as we discussed offline.
   This looks to me simplifying everything because we use all shared slots (already failed or not) to start the bulk check.
   The bulk will be just stopped immediately if any physical slot has already failed.
   If this is still too complicated or no time, I will change it to the precondition that the physical slots are not expected to fail immediately.
   


----------------------------------------------------------------
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:
users@infra.apache.org